0 calificaciones0% encontró este documento útil (0 votos)
60 vistas10 páginas
This paper was prepared for presentation at the 2008 SPE / DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 19-23April2008. Enhanced gas recovery by gas-gas displacement can be achieved economically in several situations. For mature volumetric gas reservoirs suffering from low productivity due to low reservoir pressure, injection of waste gas can increase the ultimate gas recovery. For water-driven gas reservoirs, pressure maintenance by gas injection will serve to (1) retard the influx of A
This paper was prepared for presentation at the 2008 SPE / DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 19-23April2008. Enhanced gas recovery by gas-gas displacement can be achieved economically in several situations. For mature volumetric gas reservoirs suffering from low productivity due to low reservoir pressure, injection of waste gas can increase the ultimate gas recovery. For water-driven gas reservoirs, pressure maintenance by gas injection will serve to (1) retard the influx of A
This paper was prepared for presentation at the 2008 SPE / DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 19-23April2008. Enhanced gas recovery by gas-gas displacement can be achieved economically in several situations. For mature volumetric gas reservoirs suffering from low productivity due to low reservoir pressure, injection of waste gas can increase the ultimate gas recovery. For water-driven gas reservoirs, pressure maintenance by gas injection will serve to (1) retard the influx of A
2 Sequestration by Injection of Exhaust Gases from Combustion of Bitumen Steve S. K. Sim, Alberta Research Council; Patrick Brunelle, Quadrise Canada Fuel Systems Inc.; Alex T. Turta, and Ashok K. Singhal, Alberta Research Council Copyright 2008, Society of Petroleum Engineers
This paper was prepared for presentation at the 2008 SPE/DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 1923April2008.
This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.
Abstract
Enhanced gas recovery by gas-gas displacement can be achieved economically in several situations. For mature volumetric gas reservoirs suffering from low productivity due to low reservoir pressure, injection of waste gas can increase the ultimate gas recovery by maintaining gas production rates and preventing premature well abandonment. For water-driven gas reservoirs, pressure maintenance by gas injection will serve to (1) retard the influx of aquifer and (2) partially mitigate water coning caused by excessive pressure drawdown.
This paper presents the results of laboratory core displacement tests conducted to investigate the feasibility of enhanced natural gas production by using exhaust gas from combustion of bitumen in an oxygen rich atmosphere. A synthetic gas mixture containing carbon dioxide, nitrogen and sulfur dioxide was used to represent the exhaust gas of interest. Displacement tests were conducted in Berea core and in porous media prepared with silica sand as well as crushed carbonate rocks at pressures ranging from 0.69 to 6.2 MPa. The objectives of the experiments were to determine the effects of (1) pressure, (2) displacing gas composition (3) formation water and (4) rock mineralogy on recovery efficiency of uncontaminated methane from the porous media.
Several interesting phenomena were observed during the course of this investigation. Separation of injection gas components was observed in the effluent gas during displacement. Breakthrough of carbon dioxide and sulfur dioxide were delayed relative to nitrogen. This can be attributed to the higher solubility of CO 2 and SO 2 in water relative to nitrogen. These results are beneficial to natural gas production as they reduce the operating costs associated with corrosion during production of CO 2
and SO 2 . The amount of green house gases and acid gases being sequestered in the reservoir will also increase due to these effects.
INTRODUCTION
Improved oil and gas recovery by means of carbon dioxide injection has attracted increasingly more attention in recent years due to (1) rising value of petroleum and (2) recognition that geological storage of CO 2 in depleted oil and gas reservoirs could be a potential solution to global warming by controlling the amount of green house gases in the atmosphere. Enhanced oil recovery (EOR) by miscible CO 2 flood has been practiced worldwide since the 1960s and is considered to be a mature technology. However, the idea of injecting wasted gas for enhanced gas recovery (EGR) is rarely practiced. So far, the only published field application of CO 2 injection for EGR was conducted during 1986-1994 in the Budafa Szinfelleti Field of Hungry (Papay, 1999). Here, the injected gas consisted of 80% and 20% methane from an adjacent natural CO 2 pool.
Similar to EOR with miscible CO 2 technology, EGR by gas injection can increase the ultimate recovery by prolonging hydrocarbon production rates and preventing premature well abandonment. Also, pressure maintenance by gas injection will serve to (1) retard the influx of aquifer and (2) partially mitigate water coning caused by excessive pressure drawdown. During CO 2 EOR, it is preferable to operate under miscible conditions. Mixing of CO 2 with the crude oil was encouraged since it will improve recovery by (1) reduction of oil viscosity (2) swelling of crude oil leading to reduced residual oil saturation (3) minimizing gas channeling and improved sweep efficiency. In EOR, separation of CO 2 from produced oil is 2 SPE 113468 readily achieved during depressurization. On the other hand, mixing of natural gas with the injected CO 2 or waste gases during EGR should be minimized. It will reduce the calorific value of the produced gas and render it non-marketable. Separation of CO 2 and nitrogen from natural gas requires substantial amount of capital and operational expenses.
Thus, to achieve both technical and economical success in EGR operations, contamination of natural gas by the injection gas should be minimized. A logical first step to achieve this objective would be to identify the key factors that affect the extent of mixing during gas-gas displacement within porous media. Very few research results relevant to this area (Mamora 2002, Papay 1999, Turta 2007, Singhal 2007) have been published so far.
This paper presents the results of laboratory tests conducted with two injection gas mixtures displacing methane from three types of porous media. The first injection gas mixture consisted of approximately 90% CO 2 , 5% N 2 and 5% SO 2 . This gas mixture simulated the exhaust gas produced during combustion of bituminous crude in an oxygen enriched atmosphere. The energy produced during combustion can be used for steam or power generation and the exhaust gas can be disposed into depleted oil or gas reservoirs. Due to its low nitrogen content, the density of this exhaust gas is fairly high and a large amount can be stored at the initial pressure conditions of most depleted oil and gas reservoirs. The cost of injection of this exhaust gas will be significantly lower than flue gases, which normally contain over 80% nitrogen. A second injection gas under investigation contained 60% SO 2 and 40% CO 2 . This gas mixture simulates the acid gas produced from a sour gas plant. The H 2 S present in the original acid gas was replaced by SO 2 during the laboratory test for operational safety considerations. The high concentration of SO 2 in the gas mixture allowed us to determine the limit of retention capacity for SO 2 by the connate water normally present in a gas reservoir. It is interesting to note that Sayegh et al. (1987) had conducted miscible EOR studies using similar gas mixtures. They reported that miscibility with SO 2 and CO 2 mixtures was achieved at lower minimum miscibility pressure and yielded higher oil recovery than using CO 2 or SO 2 alone.
The present gas-gas displacement tests were conducted at pressures ranging from 0.69 to 6.2 MPa. The objectives of the experiments were to determine the effects of (1) pressure, (2) displacing gas composition (3) formation water and (4) rock mineralogy on recovery efficiency of uncontaminated methane from the porous media used.
EXPERIMENTAL
The experimental set-up consisted of four main components: the porous medium, injection, production and data recording systems (Figure 1). The injection system consisted of several floating piston transfer vessels connected to a HPLC water pump. A mass flow controller regulated the gas injection rate during core flood tests. These Experiments were conducted in three types of porous media. They were (1) 30 cm x 3.8 cm diameter Berea Sandstone core, (2) 2 m long x 5 cm diameter silica sand pack, and (3) 2 m long x 5 cm diameter crushed carbonate pack. The properties of the porous media are listed in Table 1.
Produced gas volume was measured with a wet test meter and a gas chromatographic system was used for analyses of produced gas composition. System pressure was maintained by means of a backpressure regulator. A gas sampling device was installed immediately downstream of the production end of the porous medium delivering a small volume of the produced gas to the gas chromatograph at set intervals.
The porous medium was initially fully saturated with water or brine. Irreducible brine saturation was obtained by injecting methane to displace water at a slow rate and in a vertically downward direction until water production ceased. The produced water was collected in a windowed cell for volume determination and calculation of irreducible water saturation. During the gas-gas displacement test, the displacing gas was injected at a constant flow rate by means of a mass flow controller. The produced gas, after passing through the back pressure regulator, was directed to a west-test meter. Changes in produced gas composition were monitored by analyzing the gas sample received from the automatic gas sampling device at set intervals.
Test 1: Displacement of methane with synthetic exhaust gas from sand-pack at 0.69Mpa and 22 o C
During Test 1, the acid gas was injected into the sand-pack at a linear flow velocity corresponding to approximately 2 m/day. As expected, only methane was produced during the early stage of the test. Nitrogen breakthrough was observed at 20 hrs (0.65 PV injected) from the start of the test. This was followed by the first appearance of CO 2 three hours later (0.76 PV). Overall methane recovery at 2% CO 2 contamination (allowable CO 2 in sales gas) was 81% of original-gas-in-place (OGIP). No sulfur dioxide was produced during 2 pore volumes of acid gas injection. Composition of the effluent gas and methane recovery from Test 1 are presented in Figure 2.
Test 2: Displacement of methane with synthetic exhaust gas from sand-pack at 3.45 MPa and 22 o C
A second test was conducted at conditions identical to the first test except that the displacement pressure was increased from 0.69 to 3.45 MPa. As shown in Figure 3, breakthrough of nitrogen and carbon dioxide occurred at 0.73 PV and 0.96 PV injected, respectively. Significant delay in the breakthrough time of nitrogen and carbon dioxide were observed during displacement at 3.45 MPa relative to displacement at 0.69 MPa. As a result, methane recovery at 2% CO 2 contamination was increased to 97.4% OGIP. Results from Tests 1 and 2 are compared in Figure 4. It shows that the mixing zone (defined as the period in which impure methane was produced) of Test 2 was significantly narrower than that of the Test 1. These results indicate that displacement at high pressure would result in a better recovery of uncontaminated natural gas. These observations are in agreement with the results of Sigmund (1976) showing that the magnitude of molecular diffusion coefficient is lower at higher pressure. It also suggests that molecular diffusion is a dominating factor affecting the mixing of gases during gas-gas displacement.
Similar to Test 1, no sulfur dioxide was produced during 2 pore volumes of acid gas injection into the sand pack in Test 2. De-ionized water was injected into the post-test sand-pack to recover the SO 2 . The first liter of aqueous washing was very acidic giving a pH value of 1.9. This solution had a pungent odor indicating that it contained a large amount of sulfur dioxide. The pH value of the second liter of washing was 6.6 and was odorless. These results suggest that, due to the high solubility of SO 2 in water, the limited amount of injected SO 2 gas was retained by the connate water within the sand-pack during the course of the acid gas displacement. Retention of CO 2 and SO 2 due to dissolution in water is beneficial to natural gas production as they reduce the operating costs associated with corrosion during production of CO 2 and SO 2 . The amount of green house gas and acid gas being sequestered in the reservoir will also increase due to these effects.
Test 3: Displacement of methane from carbonate-pack at 0.69 MPa and 22 o C
Tests 3 and 4 were used to study the basic gas-gas displacement mechanism and fluid/rock interaction occurring in a limestone reservoir during acid gas injection. The porous medium was prepared by packing crushed dolomite (250-840 m) in a 5 cm diameter and 2 meter long stainless steel pipe.
Test 3 was conducted by displacement of methane with the synthetic exhaust gas mixture (90% CO 2 , 5% N 2 and 5% SO 2 ) at 0.69 MPa and 22 o C at a flow velocity of 2 m/day. As shown in Figure 5, breakthroughs of nitrogen and CO 2 were observed at 0.64 PV and 0.74 PV injections, respectively. Recovery of methane at 2% CO 2 contamination was 75.5% OGIP. As expected, no production of sulfur dioxide was observed during 2 pore volumes of acid gas injection. The first liter of produced water recovered from water flooding the post-test porous media was almost neutral with a pH value of 6.6. This indicates that all the sulfur dioxide injected into the carbonate pack had reacted and converted to sodium sulfite and carbon dioxide as shown in equation (1)
SO 2 (gas) + CaCO 3 Ca ++ + SO 3 -- + CO 2 (gas) ---------------------------------- (1)
Figure 6 compares the composition profile of the produced gases from Tests 1 and 3 as a function of pore volume of gas injected. It shows that the mixing zone (represented by the period in which contaminated methane was produced) of Test 3 is slightly wider than that of Test 1. It is suspected that the displacement front may have been disturbed during chemical reaction between SO 2 and the carbonate rock resulting in additional mixing that causes the widening of the mixing zone.
Test 4: Displacement of methane from carbonate-pack at 3.45 MPa and 22 o C
Test 4 was conducted at conditions identical to Test 3 except that the displacement pressure was increased from 0.69 to 3.45 MPa. The objective of this test was to investigate the effect of pressure as well as the amount of injected acid gas on the displacement process. As shown in Figure 7, breakthrough of nitrogen and CO 2 occurred at about the same time (0.67 PV injected). This is significantly earlier as compared to Test 2 conducted in silica sand at the same pressure and flow conditions. 4 SPE 113468 As a result of the early CO 2 breakthrough, recovery of marketable methane was reduced to 70% of OGIP as compared with the 97.4% obtained in Test 2. Similar to Test 3, no sulfur dioxide was produced during 2 pore volumes of injection. The pH of the aqueous washing from the post-test core sample was 6.4 indicating that the injected SO 2 had reacted with the carbonate rock. No significant change in permeability of this carbonate pack was observed as a result of the chemical reaction. Sayegh et al. (1987) had observed drastic reduction in permeability of a 4 mD Indiana limestone core plug during displacement of crude oil with a mixture of SO 2 and CO 2 . The fact that reduction in permeability was not observed in the present test could be due to the extremely high permeability of the unconsolidated carbonate pack being used.
Our laboratory observation, that chemical reaction between the injection SO 2 and the carbonate rock had resulted in an increase of mixing between displaced and displacing gas, could conceivably be just a laboratory phenomenon due to the relatively short displacement length. The disturbance generated near the well bore region may not have any significant effect on the displacement front farther away from the well bore. Further laboratory test work in this area will be required to clarify this observation.
Test 5: Displacement of methane by SO 2 and CO 2 mixture from Berea core sample at 6.2 MPa and 70 o C
The fifth test was conducted to study the recovery of methane gas from a deeper gas reservoir having temperature of 70 o C and a pressure of 6.2 MPa. A synthetic injection gas mixture containing 60% SO 2 and 40% CO 2 was used to recover methane from a 30 cm long x 3.8 cm diameter Berea sandstone core. The test results are presented in Figure 8, showing that recovery of marketable methane was approximately 80% OGIP under the test conditions. In contrast to the previous tests, production of SO 2 was observed after about 1.3 pore volumes of acid gas injection. In spite of the high solubility of SO 2 in water, the limited volume of connate water (~ 20 ml) in the Berea core was eventually saturated by the large concentration of SO 2 in the injection gas. Thereafter, SO 2 could no longer be retained within the porous medium.
NUMERICAL SIMULATION OF LABORATORY TESTS
To gain a better understanding on the interrelationship between the various factors affecting the gas-gas displacement efficiency, and to lay the ground work for future field scale modeling of the EGR process, numerical simulations were conducted to match the laboratory test results. These simulations were conducted with Computer Modelling Groups GEM compositional simulator (CMG 2007). The numerical model consisted of 1000 x 1 x 1 grids in the x- y- and z- directions, respectively. The dimension of the grid cells in the x- y- and z- directions of the model representing the 2 m sandpack were 0.2 cm , 4.5 cm and 4.5 cm respectively. The reason for using a large number of grid cells in the x-direction was to reduce numerical dispersion to a minimum (Camy, 1977). The phase behavior of the gas mixtures was modeled with the Peng Robinson Equation of state. Henrys constants of the individual gas were used to account for the gas solubility in water. Maximum gas injection rate and minimum bottom-hole pressure were applied as constraints during the simulation runs. Molecular diffusion coefficient was used as a variable to obtain a good match of the experimental results. As expected, increase in the magnitude of molecular diffusion coefficient will result in a wider mixing zone during the displacement. On the other hand, increase in water solubility of gases (i.e. decrease in Henrys constant) will result in a narrower mixing zone.
Figure 9 compares the experimental data with simulation results using the appropriate values of molecular diffusion coefficient and Henrys constants for CO 2 and SO 2 (summarized in Table 2). The match is reasonably good. In particular, the simulation model was able to reproduce the concentration profile of produced nitrogen passing through a maximum of 18% at one pore volume throughput. It should be noted that the nitrogen concentration in the injected gas mixture was only 4.93%.
Figure 10 compares the simulation and experimental results for Test 4. Although Test 4 was conducted at the same pressure and temperature as Test 2, a much higher value of molecular diffusion coefficient (0.006 cm 2 /sec for Test 4 versus 0.0002
cm 2 /sec for Test 2) was needed in order to obtain a good match. Also, to account for the conversion of SO 2 to CO 2 during displacement, a mixture containing 95% CO 2 and 5% N 2 was used as the injection gas during the simulation instead of the 90% CO 2 , 5% N 2 and 5% SO 2 that was actually injected during the experiment. The successful matching of the experimental results shows that the gas-gas displacement process can be adequately modeled by using appropriate molecular diffusion values. Table 2 summarizes the values of molecular diffusion coefficients and Henrys constants used for matching the laboratory test results.
Table 2: Input Parameters Applied in Numerical Simulation of Laboratory Tests Pressure Temp Molecular diffusion coefficient for all Henrys constant for CO 2
Henrys constant for SPE 113468 5 components SO 2
MPa o C cm 2 /sec MPa MPa Test 1 0.69 22 0.002 138 1.38 Test 2 3.45 22 0.0002 138 1.38 Test 3 0.69 22 0.004 138 1.38 Test 4 3.45 22 0.006 138 1.38 Test 5 6.2 70 0.0002 138 1.38
CONCLUSIONS
The followings conclusions were made from the results of the laboratory gas-gas displacement tests presented here:
Recovery of uncontaminated methane gas in the range of 70-90% of original gas-in-place can be obtained by linear displacement in a relatively homogeneous porous medium with an exhaust gas containing mostly carbon dioxide with low percentages of nitrogen and sulfur dioxide.
Comparison of effluent gas compositions indicates that contamination of produced methane was reduced during displacement at higher pressure. This observation is in agreement with the published results of the molecular diffusion being lower at higher pressure. It suggests that molecular diffusion is a dominating factor affecting mixing during gas-gas displacement.
Breakthrough of carbon dioxide was delayed relative to nitrogen due to its higher solubility in water. Due to extremely high solubility of SO 2 in water, all injected sulfur dioxide would be retained by the connate water within the porous medium unless the amount of SO 2 injected exceeds the retaining capacity of the interstitial water.
Reaction between sulfur dioxide and the carbonate rock was evident from the neutral pH of the produced water during the post-test water-flood. Chemical reaction within the porous medium appeared to have caused an increased mixing of displaced and displacing gases in the laboratory environment. However, its implication during field operation is not clear and requires further investigation.
Camy, J.P. and Emanuel, A.S. :Effect of Grid Size in the Compositional Simulation of CO 2 Injection, SPE paper 6894 presented at the SPE annual Fall Technical Conference and Exhibition, 9-12 October 1977, Denver Colorado.
Mamora, D.D. and Seo, J.G.: Enhanced Gas Recovery by Carbon Dioxide Sequestration in Depleted Gas Reservoirs, presented at the SPE annual Technical Conference and Exhibition, September 29-October 2, 2002, Houston,.
Papay, J,:Improved Recovery of Conventional Natural Gas, : Part I: Theoretical Discussion of Recovery Methods, Erdoel, Erdgas, Kohle,. NO 6, pp. 302-308, January 1999 and Part II: Results of a Pilot Test; Erdoel, Erdgas, Kohle, NOS 7-8, pp. 354-355, July -August 1999.
Sayegh, S.G., F.F. Krause and J.E. Fosti,:Miscible Displacement of Crude Oil by CO 2 /SO 2 Mixtures, SPE Reservoir Engineering, May 1987 pp. 199-208.
Sigmund, P. M.:Prediction of Molecular diffusion at Reservoir Conditions. Part I. Measurement and prediction of molecular diffusion co- efficients, JCPT, pp. 48-57, April-June 1976.
Singhal, Ashok K., Alexandru T. Turta, Steve S. Sim and Blaine F. Hawkins,: Enhancing Natural Gas Recovery form Nearly Depleted Gas Reservoirs by Injecting Waste Gases, presented at the 8 th Canadian International Petroleum Conference, Calgary, Alberta, June 12- 14 2007.
Turta, Alex T., Steve S. K. Sim, Ashok K. Singhal and Blaine F. Hawkins,: Basic Investigation on Enhanced Gas Recovery by Gas-Gas Displacement, presented at the 8 th Canadian International Petroleum Conference, Calgary, Alberta, June 12-14 2007.
Figure 1: Experimental Setup of Gas-gas Displacement Test
6 SPE 113468
Figure 2: Test #1- Displacing CH4 @ 0.69 MPa in sand pack 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Pore volume C o m p o s i t i o n
( % ) 0 20 40 60 80 100 0 10 20 30 40 50 60 Time (hr) R e c o v e r y
( %
O G I P ) N2 CH4 CO2 recovery Sand Pack 1 foot Berea core
BPR To Gas Chromatograph
Wet Test Meter
Mass Flow Controller Displacing gas
SPE 113468 7
Figure 3: Test 2-Displacing methane @ 3.45 MPa in sand pack 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Pore Volume C o m p o s i t i o n
( % )
0 20 40 60 80 100 0 10 20 30 40 50 Time (hr) C H 4
R e c o v e r y
( %
O G I P ) N2 C1 CO2 recovery
Fig 4: Comparison of effluent composition in Test 1 (0.69MPa) with Test 2 (3.45 MPa) 0 20 40 60 80 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Pore volume C o m p o s i t i o n
Figure 5: Test 3 - Displacing methane in carbonate pack @ 0.69 MPa 0 10 20 30 40 50 60 70 80 90 100 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Pore volume C o m p o s i t i o n
( % )
0 10 20 30 40 50 60 70 80 90 100 R e c o v e r y
( O G I P ) %N2 %CH4 %CO2 recovery
Figure 6: Comparison of Test 1 with Test 3 0 20 40 60 80 100 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Pore volume C o m p o s i t i o n
Figure 7: Test #4- Displacing methane @ 3.45 MPa in carbonate pack 0 20 40 60 80 100 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 Pore volume C o m p o s i t i o n
( % ) 0 20 40 60 80 100 C H 4
R e c o v e r y
(
%
O G I P ) % N2 % CH4 % CO2 recovery MIXING ZONE
Fig. 8 : Test 5: Displacing CH4 with 60% SO2 and 40% CO2 @ 6.2 MPa and 70 o C 0 20 40 60 80 100 0 5 10 15 20 25 Time (hrs) C o m p o s i t i o n
( % ) 0 20 40 60 80 100 C H 4
R e c o v e r y
( %
O G I P ) CH4 CO2 SO2 CH4 recovery 10 SPE 113468
Figure 9: Matching of Test #2 Displacing methane @ 3.45 MPa in sand pack 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 30 35 40 45 50 Time (hrs) C o m p o s i t i o n
Figure 10: Matching of Test #4 (injection gas contains 95%CO2 & 5% N2 during simulation) 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 30 35 40 45 50 Time (hrs) C o m p o s i t i o n