Although the basic technology for injecting carbon dioxide (CO2) for enhancing oil production has not significantly changed in recent years, methods for monitoring and controlling injection are still being refined.

Operations in the Postle field, Texas County, Okla., show how an operator uses chromatographs and remote terminal units to monitor injection in a CO2 flood.

Operations
For many years, Mobil Exploration & Producing U.S. Inc. has used "water alternating gas" (WAG) techniques in its CO2 flooded fields. This concept relies on the premise that injected CO2 mixes with oil in the reservoir, creating a lighter, easier way to move fluid.

In the WAG process, CO2 is injected for a period of time. After CO2 injection is stopped, water injection commences.

Water pushes the lightened oil toward adjacent producing well bores. This CO2-laden crude oil is pumped to the surface and then piped to a central tank battery where oil, water, and gas are separated. After separation, the water goes to a facility for reinjection, the oil is sold, and the gas is piped to a treatment plant.

In the treatment plant, the gas is dehydrated and natural gas liquids may be separated for sale. The dehydrated CO2 stream is then compressed and returned for reinjection.

The WAG process may continue for many years until the oil in the reservoir is depleted.

During secondary recovery with water injection, Mobil's Postle field had a producing capacity of 23,000 b/d. The field became a candidate for tertiary CO2 recovery when production declined to about 2,000 b/d. With the CO2 WAG technique, production is expected to reach 12,000 b/d.

Postle receives CO2 from two sources. Virtually pure CO2 is delivered by pipeline from a CO2 source field in New Mexico, and the remaining CO2 is derived from recycled gas with a CO2 concentration as low as 85%. Typically, the two streams are combined to produce a mixture varying between 93 and 97% CO2.

Because of piping restrictions, pressures at the injection wellheads deviate by as much as 200 psi from the pipeline delivery point. Process temperatures will vary from 55° F. in the winter to 75° F. in the summer, but are relatively consistent throughout the system.

The operations require precise measurements. Partners are involved in the operating expense burden, so that accurate accounting for all parties is required.

In addition, optimum operations call for verification of pattern sweep efficiency, which is the economical use of CO2. This requires affordable methods for monitoring and control at the wellheads.

Flow computation
Flow rates can be calculated with fundamental principles of fluid mechanics, shown as a set of flow equations in the American Gas Association Report No. 3, Third Edition, 1990.

Mobil selected this method primarily because of the wide range of empirical data relating to differential-producing orifice flowmeters and the corresponding correlation to wedge meters.

The volumetric flow rate at base conditions as given by Report No. 3 is:

Qv = Nl Cd Ev Y d 2 (rt,pDP)1/2 /rb

where:
Qv = Volumetric flow rate at base (standard) conditions
Nl = Unit conversion factor
Cd = Orifice plate discharge coefficient
Ev = Velocity of approach factor
Y = Expansion factor
d = Orifice plate bore diameter calculated at flowing temperature
rt,p = Fluid density at flowing conditions (based on flowing temperature and pressure)
DP = Orifice differential pressure
rb = Fluid density at base conditions.

The most accurate application of the equation dictates the measurement of three variables:

1. Flowing pressure
2. Orifice differential pressure or delta pressure
3. Flowing temperature.

All three variables can be readily monitored with transducer technologies.

Density effect
A factor of particular interest is the mass per volume, or density (rt,p, rb) of the process. CO2 fluid-stream density is greatly affected by pressure, temperature, and component mixture. This fact can be intuitively deduced when one realizes that the fluid is typically somewhere between a liquid and a gaseous state.

Leaving all other variables the same and modifying only the component concentrations to 98% CO2, 0.67% C1, 0.67% C2, and 0.67% C3 causes density to change to 55.26 lb/cu ft. That is a difference of almost 10%.

The density of a more practical injection mix of 98.5% CO2, 1% C1, 0.4% C2, and 0.1% CO3 at 1,800 psia and 72° F. is 53.01 lb/cu ft. If the mix changes to 93% CO2, 3% C1, 3% C2, and 1% CO3, and pressure drops to 1,600 psia, the density decreases to 47.88 lb/cu ft. Once again, that is a difference on the order of 10%.

To translate this example into volumetric flow terms, the following conditions can be assumed:

* Orifice diameter = 3.5 in.
* Pipe diameter = 6.065 in.
* Base conditions = 60° F., 14.65 psia
* Delta pressure = 50 in. of H2O.

The flow rate of the first mixture is 30.2 MMscfd, compared to the second of 29.3 MMscfd. Therefore, a 3% error would occur if one did not consider density changes.

Automation
Density can be measured directly with an instrument appropriately named a densitometer. From perspectives of initial capital outlays as well as ongoing maintenance expenses, the drawback is the cost of installing densitometers at multiple locations.

Therefore, Mobil sought other methods for determining CO2 density in the Postle field.

In the initial configuration, the system consisted of an RTU located at two critical measurement points. One is the pipeline delivery point for the purchased CO2 and the other is downstream of injection sites, which have different partner participation than the sites in the remainder of the field.

This placement enables equitable distribution of process costs between different entities.

The RTUs execute a real-time program to monitor the pressure, temperature, and delta pressure occurring at an orifice or wedge device. Each RTU communicates via radio links to a central computer in the field office.

A gas chromatograph at the master meter site is similarly linked to the computer. The master computer program continually extracts the gas constituent information for subsequent relay to all RTUs.

Embedded firmware in the RTUs provide precise calculations of density and heat-capacity ratio, as well as a close estimate of viscosity. Heat-capacity ratio and viscosity influence the calculation of flow rate factors such as Y and C.

Applying the AGA No. 3 flow equation with these parameters results in an accurate CO2 volume accumulation.

The gas chromatograph is housed in two industrial-style enclosures mounted near an RTU site. For ease of maintenance and to eliminate potential corrosion of electronic components resulting from exposure to combinations of CO2 and water vapor, the analyzer is separated from the controller.

The unit is a single-stream device capable of furnishing a C5+ analysis. This range permits determination of the most common elements and compounds such as CO2, N (nitrogen), C1 (methane), C2 (ethane), C3 (propane), C4 (butane), and C5 (pentane). Results are updated via radio links about every 10 min.

A central server, or computer master station, transmits the mole fractions to individual RTUs. Each RTU performs flow rate calculations and volume accumulations.

Within an RTU, data including hourly average flowing and differential pressures, hourly average temperature, and hourly volume accumulation are maintained for a period of up to 35 days.

Should a user desire, the master station can obtain this historical information on demand. The central computer also offers an operator daily summaries for purposes of closeout.

To verify chromatograph determinations with laboratory results, the master software makes a provision for the download of a sample volume rate to an RTU. This quantity sets the volume interval by which the RTU triggers an external gas sampling mechanism.

The rate proportional sample can be subsequently analyzed monthly at an offsite facility. A user then decides whether to keep the existing chromatograph data, or to edit compositions directly.

If edited, the existing flow volumes are updated with newly calculated values. This gas sample feature permits system backup in case of chromatograph failure, and satisfies requirements for installations where a sample is necessary to adhere to field operations.

Field results
Prior to system commissioning, conventional flow computers provided volume values within ±8% of the figures supplied by densitometer-based CO2 provider companies. With the current system, the volumes are within ±0.6%.

In short, the volumes match sufficiently to deem the system accurate for stream measurement purposes and for field surveillance.

Wellhead application
With this approach proven in the initial configuration, Mobil is preparing to implement this system at the wellheads.

Because RTUs are already present in a monitoring capacity using wedge meters, and the chromatograph analysis is accessible by all RTUs, no additional equipment is required.

The incremental cost difference is limited to a firmware upgrade, which is inexpensive on a per-unit basis.

Notwithstanding the obvious initial installation advantage in the Postle field, the chromatograph technique offers distinct cost savings from a maintenance point of view.

Calibration and continued operational verification of densitometers is a time-consuming task that is difficult under field conditions. This time is increased by the travel required to reach widespread sites in a typical injection scheme.

Total maintenance time increases rapidly because of the multiple number of devices in place.

When a chromatograph is used, personnel need only work with one device at a single location, and the calibration process is straightforward. Therefore, implementation of these devices can be considered more cost effective compared to densitometers when multiple points along a common pipe system are involved.