Improving Thermal Efficiency and Reducing Emissions with CO₂ Injection during Late Stage SAGD Development

Abstract

The steam assisted gravity drainage (SAGD) process requires high energy input to maintain the continuous expansion of the steam chamber for achieving high oil recovery. In the late stage of SAGD operation where the oil rate is low and the heat loss is high from a mature steam chamber, maintaining steam chamber pressure with a lower steam injection is the key to maintaining the economic oil-to-steam ratio (OSR). Both laboratory studies and field tests have demonstrated the effectiveness of adding a non-condensable gas (NCG) to the SAGD steam chamber for improving the overall thermal efficiency. In this study, a multi-well reservoir model was built based on the detailed geological description from an operating SAGD project area, which contains thick pay and top water. Grounded with the history matching of more than 20 years of production using CSS (cyclic steam stimulation) and SAGD as follow-up process, the model was applied to optimize the operating strategies for the late stage of SAGD production. The results from this study demonstrated that the co-injection of steam with CO₂ or the injection of CO₂ only has potential to improve the OSR and reduce emissions by more than 50% through the improvement in steam-saving and the storage of CO₂. The results from reservoir modeling indicate that, with the current volume of a steam chamber and an operating pressure of 4.0 MPa, about 55 sm³ of CO₂ could be sequestrated and utilized for producing 1.0 m³ of oil from this reservoir through the replacement of a steam injection with CO₂ in the late stage of SAGD operation.

1. Introduction

Among the crude oil resources discovered in the world, heavy oil accounts for more than 70% [1]. The proven onshore heavy oil resources in China are estimated to be in excess of 20 billion tons [2], which plays an important role in stabilizing the national crude oil production target. Due to the lower mobility of heavy oil under the original reservoir conditions, thermal methods based on steam injection are the major technologies for current heavy oil production. Cyclic steam stimulation (CSS), steam flooding (SF) and in situ combustion (ISC) have been successfully applied commercially in conventional to medium heavy oil reservoirs. However, steam assisted gravity drainage (SAGD) technology has become the most effective technology for the recovery of heavy oils, especially from shallow buried heavy oil reservoirs [3].

Although SAGD usually achieves a higher than 50% recovery factor of the original oil in place (OOIP) [4], its high demand in energy input and high steam quality limits its application in thin pay and deep reservoirs. It is found that only 40–50% of the total heat injected from the wellhead is effectively used for reservoir heating in most SAGD projects due to the heat loss to the adjacent formations and the energy carried out in the production fluids [5]. The SAGD projects in Canada have cumulative steam-to-oil ratios of 2.5 to 6 [6]. That is, 2.5 to 6.0 m³ of steam is required to produce 1.0 m³ of bitumen, with emissions of CO₂ approximately in the range of 320 to 850 kg if natural gas is burnt for producing steam. The thermal efficiency becomes even lower in the late stage of SAGD development.

In order to improve the thermal efficiency of SAGD, the addition of solvents (C₄–C₁₀) to the injected steam has been studied to enhance the recovery from the combined effects of thermal and solvent dissolution [7]. Depending on the injection and operating strategies for solvent addition, those technologies include expanding solvent-SAGD (ES-SAGD) [8], the solvent-assisted process (SAP) [9], and the steam alternating solvent process (SAS) [10]. A number of field tests demonstrated the effectiveness of adding a solvent into the injected steam for improving the SAGD oil production rate and oil–steam ratio (OSR) [11,12]. However, the high cost of solvents has limited their commercial application in the field; instead, the co-injection of an NCG and steam has been proposed as a more economical alternative for SAGD operations [13]. The scaled model studies indicated that adding a small amount of an NCG to the injected steam is beneficial for improving the OSR and mitigating the effect of reservoir heterogeneity [14,15,16]. The use of NCG injection to replace steam injection has been studied as the wind-down strategy for SAGD projects [17,18,19]. Meg Energy combines the co-injection of steam and natural gas together with infill well technology [20], which has been successfully applied in the Christina Lake SAGD project in Alberta, and the recovery rate and OSR have been significantly improved. The co-injection of steam and flue gas was first tested in the UTF SAGD pilot project as a successful wind-down strategy [21] for late-stage SAGD operation. A physical modeling test indicates that the timing of a flue gas’s addition could impact the behavior of the steam chamber growth and the SAGD performance [22]. The advantages of co-injecting CO₂, C6, C7 and combined C6 and CO₂ with steam have been demonstrated experimentally, with a higher recovery achieved than that of a steam-only process from the solubility of light hydrocarbons and the CO₂ diffusion into the oil [23]. The co-injection of CO₂ with steam is considered to be a viable alternative during SAGD wind-down processes to reduce steam consumption and retain a portion of the injected CO₂ in the reservoir [24]. To monitor the CO2 transportation behavior in the reservoir between the injection and production wells, δ13C–CO₂ isotopic values as an indicator are proposed and modeled based on field SAGD conditions [25]. The field test of adding N₂ to steam was also conducted [26,27] in the SAGD projects operated by the Liaohe Oilfield in China. The purpose of injecting N₂ is not only to improve the oil–steam ratio, but also to test if the injected N₂ has the ability to reduce the temperature at the top of the steam chamber, which could suppress the rising rate of the steam chamber and delay the communication with top water. The significant reduction in temperature in the top of the steam chamber was demonstrated from temperature monitoring data [27]. Waterflooding has also been used to increase recovery in heavy oil reservoirs, some of the operational issues have been studied [28].

In addition to lowering the temperature and maintaining the operating pressure in the steam chamber, the co-injection of CO₂ and steam has additional benefits: (1) lowering the viscosity of the crude oil near the steam interface due to the dissolution of CO₂ in the oil; (2) altering the rock surface wettability due to the dissolution of CO₂ in the water, which forms acidic fluids [29,30], and (3) lowering the CO₂ emissions as the result of CO₂’s storage in the reservoir.

The operation of SAGD is usually divided into four stages based on the growth of the steam chamber, including the (1) preheating, (2) rising, (3) sideway expansion and (4) confinement of the steam chamber [31]. At the late stage of SAGD operation, when the steam chamber expands to the boundary of the well pad and is in communication with the steam chambers from adjacent well pairs, the oil rate starts to fall and the OSR drops significantly. The reservoir subject for this study contains thick top water zones and has a depth of more than 600 m. Due to the risk of top water invasion, the lowering of the operating pressure that has been commonly used in most SAGD projects during the wind-down phase is not applicable. The main objectives of this study are to identify the optimal strategies that can improve the thermal efficiency at the late stage of SAGD operation and minimize the impact of top water. The novelty of this study is its objective to test the feasibility of maintaining a relatively low temperature and a high oil viscosity in the reservoir between the top of the steam chamber and the top water zones via the co-injection of steam and CO₂, which acts as a flow barrier to reduce the top water invasion rate. Considering the lager volume and high energy of the top water zones in the actual reservoir, the results from this study may underestimate the impact of top water on SAGD performance due to its confined boundary conditions in the model. However, the study has practical significance in seeking solutions to mitigate the impact of top water on SAGD operation.

2. Reservoir Simulation Studies

The Guantao Formation in Block Du 84 of the Liaohe Oilfield of Northeastern China is a heavy oil reservoir with top and bottom water zones. Commercial development was started in 1999 using vertical well CSS, with an anticipated oil recovery factor of less than 25%. To improve the ultimate oil recovery, a pilot test of SAGD as a follow-up to CSS began in 2005 using the combination of infill horizontal wells with the existing vertical wells, and commercial expansion began in 2008. The current recovery factor over the entire development area has exceeded 45% of the OOIP, with the early SAGD pilot area exceeding 65% of the OOIP [32]. Based on the detailed geological and reservoir description, a reservoir model covering the early SAGD pilot area was built and used for simulation studies. The model contains four SAGD producers and more than thirty vertical injectors. The CMG’s thermal reservoir simulator, STARS, was used to perform the modeling studies. The STARS model is widely used in the petroleum industry for simulating thermal oil recovery processes, including those of steam, hydrocarbon and non-condensable gases, as well as chemical injection processes. The production history includes 5 years of vertical well CSS production from the year 2000 to 2005, which is then followed by more than 15 years of SAGD production using the combination of horizontal infills and the existing vertical wells.

The history match of the production performance was performed to understand the current distribution of temperature, pressure and fluid saturation in the reservoir. According to the fact that the development of the steam chamber has become very mature and the SAGD operation enters into the late stage of development in this reservoir, the objective of this study is to focus on the identification of effective operating strategies that have potential to significantly cut steam requirements while maintaining a high operation efficiency. Four different operation scenarios are studied: (1) the injection of steam only at the current rate, (2) the injection of steam only at a reduced rate, (3) the co-injection of steam and CO₂ and (4) the injection of CO₂ only.

2.1. History Match of Production Performance

Based on the full-reservoir geological model, the upscaled reservoir model used for this study includes four horizontal and more than thirty vertical wells. The total number of grids for this model is 60 × 38 × 64 = 145,920. Figure 1 shows the 3D oil saturation distribution in the model, where the low oil saturation layers represent the layer of top water. In addition to considering the heterogeneity of the reservoir’s porosity, permeability and fluid saturation, a layer of higher crude oil viscosity underneath the bottom of the top water zones is also considered based on the core sample analysis. The crude oil viscosity–temperature relationship and the relative permeability of the oil phase are plotted in Figure 2.

The distribution of reservoir porosity, permeability, and fluid saturation are considered based on the geological and reservoir description. To understand the characteristics of the steam chamber growth, the pressure, and the temperature distribution in the reservoir during the different stages of development, the thermal reservoir simulator STARS was used to perform the history matching of production data from 2000 to 2020. It can be seen from Figure 3 and Figure 4 that a general agreement is achieved between the simulation and the actual performance, with an overall relative error of less than 2%. The average field steam injection rate and production liquid rates are used as input data in the model, while the oil production rate is predicted from the model. For production controls, the model sets a minimum bottom hole pressure, which is 200–300 kPa lower than the target operation pressure in the steam chamber. A minimum of a 5 °C sub-cool temperature is applied to the production well, to avoid live steam production. The maximum injection pressure is set to be 500 kPa higher than target steam chamber pressure. The field water cut and the steam chamber pressure are matched generally by adjusting the end points of the relative permeability and compressibility of the reservoir. The difference in water production between the model prediction and the actual field data is relatively larger, especially at the late stage of the production. The lower water production predicted from the model than that in the field could result from limited volume of top water due to the confined boundary in the model, as compared to the larger top water area in the field. The main purpose of the history match is to determine the basic parameters for the numerical simulation model such as the thermal and flow parameters of the reservoir’s rock and fluids. The basic reservoir and fluid parameters used in the model are summarized in

Table 1. The main reservoir and fluid parameters used in the model.

Average pay thickness	93 m
Thickness of top water layer	20 m
Average permeability	2350 mD
Average porosity	0.32
Length of horizontal well	400 m
Thermal diffusivity of rock	4.42 × 10−7 m2/s
Oil viscosity @ 40 °C	198,000 mPa·s
Oil viscosity @ 100 °C	673 mPa·s
Initial reservoir pressure at middle depth	7.0 MPa
Pressure in top water layer	6.0 MPa
Steam chamber operating pressure	4.0 MPa
Initial reservoir temperature	38 °C
Compressibility coefficient of rock	7.8 × 10−5 1/kPa
Heat Capacity of rock	2100 KJ/m3-°C
Thermal Conductivity of rock	150 KJ/m-day-°C
Thermal Conductivity of oil	11.5 KJ/m-day-°C
Thermal Conductivity of water	53.5 KJ/m-day-°C

.

Figure 5 shows typical profiles of temperature distribution in the reservoir after 5 years of vertical-well CSS operation. It can be seen that the steam chamber in the CSS stage mainly develops in a small area around the vertical wells. Since the perforated intervals of the vertical wells are relatively low to avoid early communication with top water, the height of the steam chamber around the well is low. The CSS performance was not affected by the presence of the top water layer. Thermal communication has been achieved between some wells; however, the majority of the vertical wells have shown isolated steam chambers. The dots and lines in the Figures show the well perforation locations and trajectories. The dots with arrow indicate the injection types of the wells.

According to the field development history, four horizontal infill wells were drilled in 2003 between the vertical wells with the lateral well spacing of 35 m between the vertical and horizontal wells, preparing for the pilot test of SAGD as a follow-up process using a combination of horizontal and vertical wells. Three cycles of CSS were conducted initially in the horizontal wells to establish the thermal communication with the surrounding vertical wells. After the reasonable inter-well communication was achieved and confirmed via field testing, the horizontal wells were converted into continuous production wells in 2005 while the surrounding vertical wells were selectively turned into continuous steam injection wells. Figure 6 shows the temperature profiles predicted to be in the reservoir after 5 years of SAGD operation. It can be seen that the steam chamber at this stage is extensively developed in both vertical and horizontal directions. However, the steam chamber at this stage is still far away from the bottom of the top water layer in the model.

As the steam injection progresses, the steam chamber continues to rise, getting closer and closer to the top water. The temperature distribution of the same profiles in the reservoir after 15 years of SAGD conversion is shown in Figure 7. A portion of the steam chamber at this stage has already been communicating with the top water layer. The thermal efficiency is reduced as the steam chamber is developed in the top water layer, where thermal energy is consumed without contributing to oil production.

2.2. Operating Strategies for the Late Stage of SAGD Operation

The steam injection for this reservoir started in 2000, which can be divided into two production periods: the cyclic steam injection from 2000 to 2008, and that with SAGD as a follow-up process from 2008 to now. With more than 15 years of SAGD operation in the central part of the reservoir, the steam chamber has been fully developed with oil production continuing to decline and the OSR continuing to decrease. The current operation in the steam chamber is about 4 MPa, which is lower that the pressure of the top water zone. The heat loss to the top water and overburden is high in the late stage of the SAGD operation; however, the energy stored in the reservoir is enormous due to the large volume of the steam chamber. In order to identify the most efficient development methods, four operational scenarios are studied:

Scenario 1: The injection of steam only, but at the current injection rate (about 1100 m³/d), maintaining a constant operating pressure of 4.0 MPa in the steam chamber;

Scenario 2: The injection of steam only, but with the injection rate reduced by 2/3, from the current level of 1100 m³/d to 370 m³/d, allowing the steam chamber’s pressure to fall gradually;

Scenario 3: The co-injection of steam and CO₂, keeping the same steam injection rate as Scenario 2 and adding CO₂ to stabilize the steam chamber’s pressure at 4.0 MPa;

Scenario 4: The injection of CO₂ gas only, completely replacing the steam with CO₂ to maintain the current steam chamber pressure of 4.0 MPa.

The effect of the CO₂ injection on SAGD performance can be evaluated using the following mechanisms: (1) the support of the reservoir pressure due to the non-condensable gas property of the CO₂ in the steam chamber; (2) the lowering of the average temperature in the steam chamber due to the reduced partial pressure of the steam from adding CO₂ in the chamber; (3) the reducing of the oil viscosity in the steam chamber interface due to the dissolution of CO₂ in the oil and (4) the reducing of the heat loss to the overburden due to the accumulation of CO₂ in the top of the steam chamber. As a result, the continuous SAGD operation can be supported with less steam being injected; the thermal efficiency and the OSR can be improved significantly.

The gas/liquid equilibrium K-value between the crude oil, CO₂ and water is considered in the model to simulate the dissolution, expansion and viscosity reduction of the CO₂ in the crude oil. This is calculated according to the phase equilibrium constants calculation formula K = (KV1/P) * EXP (KV4/(T-KV5)) from the CMG user manual, where KV1, KV4 and KV5 correspond to the units of P and T. The phase equilibrium constants used in the model are given in

Table 2. The K-value coefficients for the selected components.

Coefficient	Units	Water	Oil	CO2
KV1	kPa	1.19 × 107	1.89 × 106	8.62 × 108
KV4	°C	−3816.44	−4680.46	−3103.39
KV5	°C	−227.02	−132.05	−272.99

.

A comparison of the production performances predicted from the above-mentioned four operating scenarios during the 5.0 years of the late stages of SAGD operation is shown in

Table 3. A comparison of the production performances of different injection scenarios.

Injection Scenarios	Cumulative Oil Produced (m3)	Cumulative CO2 Injection (sm3)	Cumulative CO2 Production (sm3)	Cumulative Steam Injection (m3)	Oil–Steam Ratio (m3/m3)	Net CO2 to Oil Ratio (m3/m3)	Equivalent EOSR (m3/m3)
Scenario 1	490,505	0	0	1,970,844	0.25	0	0.25
Scenario 2	316,202	0	0	678,898	0.47	0	0.47
Scenario 3	378,647	5,583,541	218,331	678,898	0.51	14.2	0.50
Scenario 4	297,804	17,772,264	1,644,599	0		54.1	1.39

and Figure 8. It can be seen that the total oil production for the same production duration is lower with the reduced steam injection rate or with the addition of CO₂, with comparison to the case with a higher steam injection rate (Scenario 1). However, the OSR is improved significantly with the reduced steam injection rate. With the steam injection rate reduced, the total enthalpy injected into the reservoir is not sufficient to support the continuing growth of the steam chamber. This will result in a lowering of the pressure and temperature in the steam chamber, which will reduce the gravity drainage rate due to an increased oil viscosity. However, heat loss is also reduced with a lower temperature and pressure in the steam chamber. Scenario 2, with a reduced steam injection rate, is not recommended, due to the risk of top water invasion due to the reduced operating pressure in the steam chamber. In Scenario 3, where CO₂ is co-injected with a reduced amount of steam at the same time, the pressure in the steam chamber can be maintained, which provides support to the continuing growth of the steam chamber. Comparing this to Scenario 2, the benefit of adding CO₂ to the reduced steam injection is clearly demonstrated, with a higher oil production achieved (see Figure 8). In Scenario 4, where only CO₂ is injected, although the steam chamber’s pressure is maintained at a constant of 4.0 MPa, oil production, however, is lower due to no additional thermal energy being added.

Although the effect of non-condensable gas on SAGD oil rates is still not clearly evidenced from field performance [20], the modeling results in Scenarios 3 and 4 indicate that the oil production is significantly lower than that of the normal steam-only injection case (Scenario 1), particularly in the late stage of operation. The reservoir model may require further validation from laboratory and field data in the future to more properly model the behavior of CO₂ gas in the steam chamber and in the reservoir. Future optimization is required for the operating parameters, such as the ratio of steam and CO₂ injection and the operating pressure, etc., to achieve higher oil rates and a higher OSR.

With a reduced steam injection or a co-injection with CO₂, the economic benefit can be evaluated from the modeling results in Table 3. Compared to Scenario 1, Scenario 2 produced 17,303 m³ less oil, but with steam savings of 1,291,946 m³. This indicates that an incremental OSR is only 0.134 if a normal steam injection rate is applied. This is a clear indication that the saved steam is more beneficial to be injected into newer SAGD wells in this reservoir where the average OSR is 0.25.

To facilitate the comparison of the thermal efficiency for the above four scenarios, an equivalent oil-to-steam ratio (EOSR) is introduced to include the CO₂ injection volume for the calculation of the total equivalent steam amount. By assuming that 75.0 sm³ of natural gas is required to generate 1.0 m³ of steam based on the field data, and assuming that the cost of CO₂ is the same as that of natural gas, the calculated EOSR is given in the last column in Table 3. It can be seen that Scenario 4, with the CO₂-only injection, achieves the highest thermal efficiency (EOSR = 1.39).

As the target formation contains top water, maintaining a relatively high pressure in the steam chamber is beneficial for balancing the pressure in the top water layer and mitigating the risk of premature communication with top water. Therefore, Scenario 2, with a reduced steam injection rate only, is not practical to maintain SAGD operations, as the falling pressure in the steam chamber could trigger an inflow of top water. In order to save steam during the late stage of SAGD operation in the reservoir with top water, a co-injection with CO₂ or an injection of CO₂ only is required to support the operation pressure in the steam chamber.

During SAGD operations, the majority of the CO₂ injected will be stored in the reservoir as long as a proper sub-cool temperature is controlled in the production wells. The amount of CO₂ produced with oil and water is relatively small due to the low solubility of CO₂ at high temperatures, which is not expected to cause much more difficulty in the treatment of the produced fluids in the surface plant’s facilities.

2.3. Effect of CO₂ on the Temperature of the Reservoir’s Top

Figure 9 shows the change in the reservoir temperature in the top layer of the reservoir model before and after CO₂ injection for 5 years. It can be seen that the temperature has reduced after CO₂ injection. Figure 10 shows that the temperature changes with time for the selected blocks at the top layer of the model. It indicates that the temperature in the top layer was not changed for the initial years of operation and then increased gradually due to the heat conduction from the steam chamber below. The temperature finally reached the value of the saturated steam; however, it started to fall after the addition of CO₂ into the injected steam and this could be due to the accumulation of CO₂ at the top of the reservoir. The accumulation of the CO₂ gas in the top of the reservoir reduces the temperature and steam requirements for the process.

It can be seen in

Table 4. Estimations of CO₂ emissions from different operating strategies.

Injection Scenarios	Oil–Steam Ratio (m3/m3)	Storage of CO2 in Reservoir (kg)	CO2 Emission from Steam Generation (kg)	Total CO2 Emission (kg)
Scenario 1	0.25	0	568	568
Scenario 2	0.47	0	302	302
Scenario 3	0.51	27	278	251
Scenario 4		101	0	(−101)

that with the co-injection of CO₂ and steam, the amount of CO₂ emissions per m³ of oil production can be reduced by more than 50%. With the complete replacement of steam with a CO₂ injection in the late stage of SAGD operations, a negative amount of CO₂ emissions can be realized, which can be used to offset the amount of emissions produced before.

3. Emission Reduction from the Injection of CO₂

The emission reduction from the co-injection of steam and CO₂ or the injection of CO₂ only can be evaluated from (1) the reduced CO₂ emissions due to the reduced fuel consumption from steam saving, and (2) the storage of CO₂ in the reservoir during CO₂ injection. To calculate the CO₂ from steam generation, natural gas is assumed to be used as the fuel in the following analysis. Based on the middle-range heating value of natural gas [33], about 1.9 kg of CO₂ is produced for every 1.0 sm³ of natural gas burnt as fuel. Therefore, about 142 kg of CO₂ is produced for every 1.0 m³ of steam generated, based on the average of 75 sm³ of natural gas being consumed in the field. About 560 kg of CO₂ emissions were reported for producing 1.0 m³ of oil from SAGD projects in Canada with a typical SOR of 2.5 [6]. In other words, 140 kg of CO₂ emissions can be reduced for every 1.0 m³ of steam saved via CO₂ injection during SAGD operation. The additional reduction in CO₂ emissions is from the storage of injected CO₂ in the reservoir, to replace the volume vacated from oil production and the condensation of steam. In this analysis, the total CO₂ emission reduction is calculated based on the CO₂ reduction from the steam being saved and the CO₂ being stored in the reservoir. The storage of CO₂ in the reservoir is the difference between the injection and the produced amounts of CO₂. Without the consideration of the emissions produced to separate and compress the CO₂ gas, the addition of CO₂ to the steam could reduce the emission amount of CO₂ significantly due to the improved OSR and the storage of CO₂ in the reservoir. Table 4 summarizes the calculated CO₂ emissions from different operating strategies based on the performance predictions from reservoir modeling studies. The amount of CO₂ reported in Table 4 is based on 1.0 m³ of oil being produced. The storage of CO₂ in the reservoir is the net amount between the injected and the produced amounts of CO₂ from the simulation results. The CO₂ emissions from the steam generation are calculated from the OSR predicted from simulation studies. It can be seen that with the addition of CO₂ to the injected steam, the amount of CO₂ emissions per m³ of oil produced is reduced significantly.

4. Conclusions

The following conclusion has been reached based on the simulation studies for the Liaohe SAGD projects.

• At the current stage of Liaohe SAGD operations, the oil rates and OSR are expected to fall continuously. The project economics may suffer with continuous high steam injection rates.
• Modeling results indicate that the OSR can be improved by 88% with the steam injection rate reduced by 66%; however, the lowering of the operating pressure in the steam chamber in this reservoir is not practical due to the high risk of top water invasion.
• By adding CO₂ into the injected steam during SAGD operations, the operating pressure can be maintained with a reduced steam injection rate. The emissions of CO₂ could be reduced by more than 50% due to the steam being saved and storage of CO₂ in the reservoir, based on the simulation results.
• The accumulation of CO₂ at the top of steam chamber lowers the temperature in the formation near the top water layer by more than 20%, which is also beneficial to mitigating the impact of top water on SAGD performance.
• With a complete replacement of steam injection with CO₂ gas during the late stage of SAGD operation, a negative emission can be achieved by effectively utilizing the remaining thermal energy in the steam chamber.

In summary, in the late stage of SAGD operations, where the steam chamber is large and oil drainage efficiency is low, continuing steam-only injection to maintain the chamber’s pressure is not cost effective. A steam–gas co-injection or a gas-only injection appears to be very effective in supporting the gravity drainage process while utilizing the enormous amount of remaining thermal energy in the mature steam chamber. The results from this study can also be applied to other SAGD or steam flood projects.