6 CO2 Plugin

	This functionality requires the CO₂ license feature

The Migri CO₂plugin has been available for Migri since release 2.0.1. It is, however, not recommended to use older releases of Migri than 2.1.0 for simulations. A CO₂ plugin license must be acquired and activated before creating or opening a CO₂injection project (see Figure 6.1).

Figure 6.1. Activate the Migri CO₂ plugin by the “Use CO2” option. The Migrisk option is not required unless you plan to do MonteCarlo CO₂ injection modelling.

The following tasks are required to use the CO₂plugin:

1. Load or build a Migri project.

2. Modify the timesteps

3. Modify the HC table

4. Edit the CO₂Injection panel

5. Simulate the CO₂ injection timestep(s)

6. Analyze the results

Load or build a Migri project.

This task involves the same steps as for building a Migri project for petroleum migration modelling. The minimum requirement is to use a single grid representing the depth to the top of the layer where CO₂will be injected. The grid-file(s) can be dragged and dropped into Migri to create a single layer 3D model. Loading the model into Migri should result in a display like the one in Figure 6.2. Note that the time annotation in the 3Dview window will be shown as “0.00 Ma” when it is a normal Migri project, while the CO₂ injection models will be annotated as “0.00 a”. Migri uses “a” for anno, which is to be interpreted as “years before the end of the simulation run”.

Figure 6.2. A 3D Migri model ready for CO₂ plugin modifications.

Modify the timesteps

Having loaded the Migri Model, the time steps to be used for simulation must be set using the Timesteps Editor. In our example, the simulation timesteps are between 0 and 50 years. if the heading of the second column is showing Times [Ma] use the selector to the right of the column heading to change it into Times [a] (as shown in Figure 6.3). Use Close to close the panel and use these timesteps. Click on the now red reload model button to “Update timestep” before continuing. This will save and reload the model with the correct timestep settings.

Please note that in release 2.1, the times are given as “time before the end of the simulation”. We may allow for “time from the present” in a later release.

Figure 6.3. The timestep editor is used to define the timesteps to be used in the modelling. Note that Times are set in years (a=anno). This simulation history starts 50 years before the end and the first timestep is from 50 to 40 years before the end

Modify the HC table

The HC Table defines the components to be used in the modelling. The CO₂ plugin requires a 2-compound mixture, where only the first compound (first row in Figure 6.4) will be of importance. The second row in the table is included because Migri requires at least 2-compounds for the PVT computations. Figure 6.4 shows that we model the system with (trace amounts of) C20 (N=20) in row2 while the CO₂ in row 1 is given the carbon number of -1 (N=-1). The Gas and Oil fractions columns of the table are not used when modelling CO₂ injection.

Figure 6.4. Example of the content of HC Table used to define the CO₂ fluid system. Only the 2 leftmost columns are used by the CO₂ plugin.

Edit the CO₂ Injection panel

The CO₂ Injection panel is where we activate the CO₂ injection modelling (see general settings in Figure 6.5). It is also the place where we define the injection history (upper table in Figure 6.5) and the injection points (lower table in Figure 6.5).

For each row in the Injection history table (Figure 6.5) time and injection values are specified. The time column values will typically be the same as the simulation times (compare Figures 6.3 and 6.5). The injection rate units supported by release 2.1 include [Mton], [Mton/year] , [cum/year] and [cum]. When a volumetric unit is used, the Injection density (default 450 kg/m³) is used to convert from volumes to mass when required. The CO₂ density (default 425 kg/m³) is used in the modelling of CO₂ injection process. We recommend using the Mton or Mton/year units as this is frequently used to document CO₂project plans.

The Injection points table at the bottom of Figure 6.5 contains the list of the starting points for the injection modelling. The points can be loaded from a csv file, or you can add picked points in the 3D model. To add new picked points with a point, close the CO₂injection panel if it is open (1) pick points in 3Dview or 2Dview (2), then open the CO₂ injection panel and use “Add injection points/picked points” button to add the picked points to the injection table. Points can also be added from a csv file using “Add injection points/File…”. Migri will assign the points to a layer number, but please do check that these are as wanted. You will find the layer number in upper left corner of Project Workspace. Migri will use the layer number (not the depth value) when determining where to inject the CO₂.

Figure 6.5 CO₂ Injection panel with General settings, Injection history with unit selections and Injection points.

Figures 6.6 and 6.7 show how to simply append a picked point to the injection points table. The appended point in row 5 is given a Layer # and a default scaling factor of 1. Here the appended point was picked on layer 5 (see Figure 6.7), and we would want change it to layer 17 before continuing.

Figure 6.6 Snapshot after picking a point in 3D vew and using Add Injection points/Picked points

Figure 6.7 snapshot of Injection points table after picked points have been appended in Figure 6.6

Migri will combine the Injection history table and the Injection point table when modelling the injection of Co2 for each timestep. If the data in Figure 6.5 is used, then the first and third points will be injecting 20 Mton for their first 10-year step while the 2^nd will inject 10*20=200 and the 4^th point will be injecting 200 times 20 (=4000 Mton – a quite unrealisic value…). Note that Migri uses the Easting, Northing Layer# to identify the node points where injection starts.

The CO₂ Injection settings panel will be applied and closed when pressing the Apply button at the bottom.

Simulate the CO₂ injection timestep(s)

Once the CO₂ injection settings have been completed, Migri will simulate the CO₂ migration using the normal buttons. We recommend storing BasicOutput, HC.flowrates, Retained.HCColumn (see more in Analyze the results below). The Output Editor is used to define what should be stored from the simulations.

It is recommended to use the Simulate Next Step button (see Figure 6.8) and view the results (see below) for each timestep simulated.

Figure 6.8. The Simulate toolbar contains the Simulate next step button (marked red in snapshot).

Analyze the results

After each step has been simulated Migri will usually plot the HC.flowrates. In our case the flowrates plotted are those ofCO₂ and not hydrocarbons, off course. In our demo case we show the flowrates after the first timestep in Figure 6.9. More details about the flow are shown in Figures 6.10 and 6.11 which show how both the modelled columns, saturations and velocities decrease away from the injection site.

Note that the modelled CO₂ column heights and velocities may be modelled to be much larger than in this demo case. This will depend on the geological (flow) properties used and the injection rates used. Migration velocities of 100.000 km/My corresponds to 100 m/year. the injected volumes are placed in the center of each mesh element and the near well CO₂ plume behaviour is not modelled in this release of Migri. Therefore, the migration front may be broader than in e.g. this demo example. When the columns are large, the injection may be spread across more injection point to better simulate the near-well plume spreading of the CO₂.

Figure 6.9. CO₂ flow-rates with injection starting from 3 points close to the Longhorn 31/5-7 well and one injection point to the east. Close-up of flow to the right. The Field outline in the north (white line) is that of Troll West.

Figure 6.10 Modelled retainedCO₂ column (m) after first timestep

Figure 6.11 Modelled retained CO₂ average saturation (m) after first timestep and velocity (km/My). Note that km/My is the same as 0.001 m/y. Maximum velocity is here modelled to be 18 m/y.

Further work

The 2.1.0 release of Migri contained the first release of the CO₂ plugin where this user documentation was included. It contains all the user elements needed to set up and simulate a case. Both deterministic and stochastic simulations can be run with this release. Remaining tasks to be included in later releases include:

Times to refer to year from the present; some time annotations have been improved but more to do here.
Automatic definition of the HC Table content. No need for users to be concerned with this information in later release
Renaming of e.g. HC.flowrates to CO₂ flowrates etc. Rates of migration are much higher than in PSA modelling, so a new legend will be needed for some of the plots, e.g. migration velocities.

User comments and reports of issues with the CO₂plugin and/or documentation are welcome.