SYSTEM FOR INJECTING FLUE GAS TO A SUBTERRANEAN FORMATION

Abstract

A system (100) for injecting flue gas to a subterranean formation, wherein the system (100) is configured to receive an initial mixture of N.sub.2, CO.sub.2 and less than 2% other components and comprises a compressor (110) for obtaining and maintaining a predetermined downhole pressure. The system (100) has a control system (200) for maintaining the amount of CO.sub.2 in an injection mixture in the range 12-90% and can be configured for EOR.

Claims

1. A system for injecting flue gas to a subterranean formation, wherein the system is configured to receive an initial mixture of N.sub.2, CO.sub.2 and less than 2% other components, the system comprising: a compressor for obtaining and maintaining a predetermined downhole pressures; and a control system for maintaining the amount of CO.sub.2 in an injection mixture in the range 12-90%.

2. The system according to claim 1, wherein the amount of CO.sub.2 in the injection mixture is maintained in the range 20-90%.

3. The system according to claim 1, wherein the compressor and injection mixture are configured for enhanced oil recovery.

4. The system according to claim 1, wherein the control system comprises a membrane for reducing the amount of N.sub.2.

5. The system according to claim 1, wherein the control system comprises a mixer for adding CO.sub.2 to the initial mixture.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The invention will be described in greater detail by means of exemplary embodiments and reference to the accompanying drawings, in which:

[0020] FIG. 1 illustrates a system according to the invention;

[0021] FIG. 2 illustrates efficiency at different injection rates; and

[0022] FIG. 3 illustrates efficiency at different CO.sub.2 concentrations.

DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT

[0023] FIG. 1 is a schematic illustration of a system 100 according to the invention. As indicated in the introduction, a flue gas from any combustion can be treated to provide a suitable gas mixture for the present invention. At the system boundary 1, a treated flue gas containing N.sub.2, <12% CO.sub.2 and <2% other components is assumed. For example, a combined cycle may be present upstream on a main feed 101 as described. Similarly, an afterburner or other device (not shown) to reduce the O.sub.2 content may be present upstream on a supply line 120 for adding CO.sub.2 from an external source. Downstream from the system boundary 2, i.e. to the right in FIG. 1, a pipeline conveys a compressed injection mixture to a subterranean formation.

[0024] The system 100 comprises a control system 200 for controlling the composition of the injecting mixture, which is compressed to a desired pressure by a compressor 110. The system 100 may comprise other parts, e.g. an intercooler 120. The intercooler 120 is a commercially available, standard system component in many compression systems.

[0025] The control system 200 comprises a membrane 210 for separating N.sub.2 and a mixer 220, e.g. a controllable valve. A sensor 230 is shown downstream from the membrane 210 and mixer 220 to illustrate a feedback loop. The sensor 230 may alternatively be disposed upstream to implement a feed forward loop. Either way, a controller 240 receives input from the sensor 230 and provides a response to an actuator, in FIG. 1 represented by the mixer 220. The controller 240 comprise hardware and software to execute a cybernetic algorithm, e.g. a feedback or feed forward algorithm. The controller and algorithms are known to one skilled in the art.

[0026] In the following, we use measured values from a combined cycle as a numerical example. In particular, the initial flue gas from a typical gas turbine contains 5% CO.sub.2, 74% N.sub.2, 15.5% O.sub.2 and 5.5% other components. This O.sub.2 content is too high for EOR applications. A secondary step involving a steam generator and a steam turbine provides a reference flue gas containing 11.4% CO.sub.2, 86.9% N.sub.2, 1% Ar, 0.6% O.sub.2 and 0.03% H.sub.2O.

[0027] This mixture can be passed through a commercially available filter in order to reduce the content of N.sub.2. A numerical example is provided in table 1, which is computed from the mixture above using an Aspen Process Simulation System, provided by Air Products Ltd. (www.airproducts.com), with a PA405N1 membrane model.

TABLE-US-00001 TABLE 1 Membrane filtering of reference flue gas N.sub.2 O.sub.2 CO.sub.2 H.sub.2O Ar Other Released   49% 0.2% 0.4% — — 0.6% Deposited 36.3% 1.7% 10.8% — 1.25% —

[0028] The row ‘Released’ contain fractions released to the atmosphere, and the row ‘Deposited’ contains the components that do not pass the membrane, and thus are eligible for injection. Disregarding the fractions released to the atmosphere and noting that the fraction in the ‘Deposited’ row add to about 50%, it is readily seen that the ‘Deposit’ fraction or injection mixture contains about 72.6% N.sub.2, 3.4% O.sub.2, 21.5% CO.sub.2 and 2.5% Ar. The value provided for Ar should be interpreted as the fraction of ‘other components’, e.g. NOx.

[0029] An alternative to membrane filtering is to add CO.sub.2 from some external source to achieve a fraction of CO.sub.2 above 12%, preferably above 20%, in the injection mixture.

[0030] Several alternatives for EOR using flue gas as injection fluid have been compared to a base line using water as injection fluid. More particularly, The Eclipse 300 2013.2 software was used for EOR simulations and the Eclipse PVTi 2013 package was applied for the associated PVT models. First, the baseline was established using 5000 m.sup.3 at 58 kg/s water injection. Next, flue gas injections was simulated using different gas mixtures and alternating gas injection with water injection. The ‘other components’ were treated as N.sub.2 in the simulations.

[0031] The following assumptions, corresponding to sandstone, were made for the reservoir:

Porosity: from 15% to 25%, mean=19%
Permeability: 160 to 650, mean=385 mD

Perm Z=(Perm X)*0.5

[0032] Netto-gross: 0.56 to 0.76 (net formation thickness contributing to oil and gas production/gross thickness of formation)
Bottom of well pressure: 68 bars+Δ10 bars
Oil production: 5000 m.sup.3/day.

[0033] FIG. 2 illustrates the oil recovery rate or efficiency as a function of time (15 years). The baseline 10 was obtained using water at a flow rate of 58 kg/s. The efficiency after 15 years is 28%. The curve 20 represents an alternative using flue gas directly, and was obtained using the reference flue gas containing 11.4% CO.sub.2 at a flow rate of 40 kg/s. The efficiency after 15 years is 20.5%. Curve 22 was obtained using the reference flue gas, i.e. as for curve 20, but at an increased flow rate of 80 kg/s. This case is actually better up to year 2, but then a breakthrough from injection well to production well caused a large area with poor sweeping effects between the wells, so that gas injection became circulation.

[0034] FIG. 3 illustrates the oil recovery rate or efficiency as a function of time (15 years, commencing in 2015), i.e. as in FIG. 2. The baseline 10 represents water injection at 58 kg/s as in FIG. 2. Curve 30 represents a preferred alternative, i.e. a flue gas with increased CO.sub.2 content, and was obtained using the injection mixture with 21.5% CO.sub.2 specified above at a mass flow rate 17 kg/s. Up to year 11 (2028), the oil recovery rate is identical. The efficiency after 15 years is approximately 27%.

[0035] From FIG. 3, it appears that an injection mixture of N.sub.2 and CO.sub.2, where the CO.sub.2 fraction is above 20 mole %, has the same EOR effects as water injection, even with flow rates at ⅓ of the water injection rate.

[0036] The above results are generally due to the properties of N.sub.2—CO.sub.2 mixtures in the range 12% to 90%, in particular to the PVT-properties or phase diagrams. Thus, they may be applicable in other compression applications, e.g. depositing CO.sub.2 in aquifers or other subterranean formations.