Michigan Basin Site - Validation Phase

Michigan Basin Site - Validation Phase

During Phase II, the MRCSP conducted two rounds of injection in Otsego County, Michigan, at a site located at the northern rim of the Michigan Basin. This portion of the basin is in an area of active enhanced oil recovery projects and natural gas processing plants. CO₂, which is a byproduct of natural gas production, periodically is used for enhanced oil recovery in the Niagaran Reefs. The host for this test was DTE Energy, which owned the gas processing plant that supplied the CO₂ for the test. The operator and lease holder of the test well site was Core Energy.

The research team applied for and received a Class V Experimental Technology, Underground Injection Control permit from US EPA Region 5 in 2007. The research team was able to test innovative monitoring methods including repeated cross-well seismic tests. A total of 60,000 tons of CO₂ was injected in saline formations (known as the Bass Islands Dolomite) located adjacent to the Niagaran Reefs in an enhanced oil recovery field operated by Core Energy in the vicinity of a DTE natural gas processing plant.  

The Michigan Basin Validation Phase field project was initiated during the fall of 2006 and completed during the fall of 2009. During that time several project snapshots were developed to provide updates on the project. The snapshots are contained below. In addition, additional information about the Michigan field test, including the project fact sheet, briefing, and fact sheet on monitoring techniques can be accessed by clicking on the links.

Project Snapshots

Summer 2009

In late fall 2009, MRCSP completed the second round of injection at the Michigan Basin site and has begun post-injection monitoring activities. This collection of snapshots shows the progression of activities from site preparation through completion of the injection. 

1. January 2009 – February 25, 2009: Well Preparation, Permitting, and Pre-Injection Testing and Monitoring

Pre-injection activities included working on the well to prepare it for injection, coordinating with U.S. EPA Region 5 to ensure compliance with the injection permit, and conducting several tests aimed to understand baseline conditions prior to continuing injection. The collected data were of good quality and showed a stable setting. Specific pre-injection activities included the following:
  • Health and Safety/Mechanical Integrity Testing. Initial pressure testing and monitoring to locate any leaks in the wellhead, packer, tubing, and pipeline. 
  • Well Indicator Sensors. Pressure and temperature sensors were installed in the tubing within both the injection and monitoring well to determine baseline conditions in the reservoir. 
  • Reservoir Brine Sampling. Brine samples were analyzed for major ions, cations and physical parameters.
  • Wireline. Wireline logs were run in the injection well and monitoring well prior to injection. A wireline test includes any test using instrumentation that is lowered into the well on a cable, or wireline. A pulsed neutron capture log, which is used to determine carbon dioxide (CO₂) location at the well bores, was one of the wireline tests performed.
  • Cross-Well Seismic Survey. This survey was used to obtain high resolution images between the injection well and monitoring well. This survey will be repeated to create time-lapsed images of the CO₂ as it moves away from the injection point. 
2. February 25, 2009: MRCSP receives written approval from U.S. EPA Region 5 to begin injection.

Good news! MRCSP received written authority from U.S. EPA Region 5 to inject and turned the system on the next morning.  

3. February 26, 2009 – March 5, 2009: Injection proceeds smoothly – 2,772 metric tons injected thus far.
4. March 6, 2009 – March 15, 2009: “Pressure Fall-Off” testing takes place.

The pressure fall off test is a procedure whereby the injection is stopped and changes in the pressure and temperature are monitored. By watching the pressure and temperature recovery in the formation, MRCSP evaluated some injection characteristics such as localized wellbore skin effects, far-field boundary affects, and any changes to permeability. Measurements taken at the bottom of the well were compared to those taken at the top. The downhole data showed that reservoir pressures were behaving normally and were consistent with values observed in last year’s test. At the end of the ten day shut-in, the monitoring well was sampled and brine samples were also collected from the monitoring well. After completing the pressure fall-off test, injection resumed.
The Service Rig
5. March 16, 2009 – April 25, 2009: Injection continues smoothly.

24,372 metric tons were injected by April 24, 2009. The MRCSP partners toured the site during this timeframe.

6. April 26, 2009 – May 19, 2009: Mid-Injection Monitoring

Injection was shut down at the end of April to allow for wireline testing and other monitoring at the mid-point of the injection. Pressure fall off was monitored for 10 days after injection stopped. Downhole pressure gauges were also collected from the injection well and the monitoring well. A pulsed neutron capture log was also completed to detect the presence of CO₂ in the monitoring well. The injection phase monitoring scheme is aimed at ensuring that the system is operating properly and uses a combination of system, health and safety, and reservoir monitoring, including the following: 
  • Injection System Monitoring. This included continuous recording of pressure, temperature, and flow rates at the pipeline and wellhead. These parameters were recorded by a data logger in approximately 15 second intervals. In addition, a sample was collected from the compression facility and analyzed in June to verify the chemical composition of the injection stream.
  • Health and Safety Monitoring. An array of gas monitoring sensors installed at the injection wellhead during the injection period were programmed to sound an alarm if higher than normal conditions exist due to some sort of large CO₂ release from the injection system.
  • Well Indicator Sensors. Pressure and temperature sensors were installed in the injection well and the monitoring well. Retrievable, downhole loggers were set to record pressure and temperature at 1 minute intervals.
7. May 20, 2009 – July 8, 2009: Injection resumed and stopped when a total of approximately 50,000 metric tons were injected.
8. July 8, 2009: Pressure fall-off test was started and will be followed by post-injection monitoring.

The objective of the post-injection monitoring is to measure reservoir conditions as they return to normal. This effort will demonstrate that the CO₂ is safely sequestered and the system may be closed. It is also expected to provide information on the sequestration mechanisms in the Bass Islands Dolomite. This work is currently ongoing.
  • Injection System Monitoring. Post-injection system monitoring will include recording of pressure and temperature in the injection and monitoring well as they decline. 
  • Well Indicator Sensors. Pressure and temperature sensors will be measured in the injection well and the monitoring well. Reservoir modeling indicates that pressures in the injection formation decline to near pre-injection levels within 2 to 3 weeks after injection is stopped.  
  • Wireline. Wireline logs will be run in the injection well and monitoring well after injection. A pulsed neutron capture log will be used to determine CO₂ location at the well bores and samples of the casing and cement will be collected for analysis.
  • Reservoir Brine Sampling. A brine sample will be collected from the monitoring well to provide an idea of the long-term changes to the formation fluid that occurred due to injection of CO₂. 
  • Cross-Well Seismic Survey. The repeat cross-well seismic survey will be completed after the pressures in the injection well and monitoring well have stabilized. The differencing of the two surveys will help show where the injected CO₂ is between the two wells. This will allow the determination of the preferential pathways of CO₂ within the injection interval.
The Service Rig (left) and Cross-Well Seismic Truck Injection Well During Routine Operation (right)
Winter 2009


In an effort to further develop our understanding of carbon dioxide (CO₂) injection into deep saline formations, the MRCSP is conducting a second, somewhat larger injection test into the Bass Islands Dolomite, which is the same saline formation used for the previous test in early 2008 at the Michigan Basin field test location. This test will involve injection of up to an additional 50,000 tonnes. The test will be conducted in compliance with all state and federal permitting requirements.

This expanded injection test and monitoring program have specific objectives that include:
  1. Documenting the movement of carbon dioxide from the point of injection past the monitoring well
  2. Confirming and refining the crosswell seismic results obtained in the first test by using a larger data set,
  3. Evaluating long term pressure and temperature responses,
  4. Continuing to monitor geochemical changes in the reservoir, and,
  5. Refining the computer model developed to predict the behavior of injected carbon dioxide in this specific formation
As indicated in the fall 2008 Snapshot, the first test at the Michigan basin site involved injection of 10,241 tonnes of CO₂. The MRCSP research team carried out a number of sophisticated measurements around the injection site to track the behavior of the CO₂ injected into the formation. These measurements showed a reasonable match with the behavior predicted by the computer model used by the MRCSP team prior to the field test. Improving and validating our ability to predict the behavior of CO₂ injected into deep geologic formations is an important step in building certainty in the integrity of carbon sequestration as a means of addressing climate change. In order to better understand CO₂ storage in deep saline formations at a larger scale, the extended test with up to 50,000 additional tonnes at our Michigan Basin site is an important next step. The larger scale and longer injection test will provide MRCSP researchers with a unique opportunity to evaluate how supercritical carbon dioxide moves through a geologic system. The total amount of CO₂ injected over the original test and the extended test of about 60,000 tonnes will be the largest deep saline reservoir injection in the United States to date and a significant step towards evaluating larger scale injection scenarios.


The project team will make minor adjustments to the injection well used for the first test and document pre-injection conditions; this step will culminate in the completion of a Mechanical Integrity Test (MIT) that is used to verify the integrity of the well and gain written approval from US EPA to begin the injection test. It is expected that this step will be completed so that injection will begin in February, 2009. The current plan, based on the results from the first test, are to inject as much as 600 tonnes of carbon dioxide per day for a period of three to nine months and for a total of not more than 50,000 metric tonnes. Once injection has ceased, the team will conduct post-injection monitoring.
Monitoring Plan

Prior to the beginning of injection, monitoring, mitigation and verification (MMV) activities take place to establish a new baseline against which future work can be compared. Samples of the brine from the monitoring well are collected and analyzed and a crosswell seismic survey is conducted between the injection well and two nearby monitoring wells. This survey creates an acoustical picture between the wells, almost like an ultrasound. This allows scientists to image the injected carbon dioxide between the wells. 

During the injection period, the monitoring activities are designed to ensure that the system is operating properly. The monitoring system will employ a combination of system, health and safety, and reservoir monitoring technologies. Sophisticated sensors that can be dropped into the monitoring wells and the injection well (called wireline tools) are being used to collect information about the behavior of carbon dioxide right next to the well. Samples of the brine from the monitoring well are collected and analyzed and compared with the baseline samples to help determine the rock’s response to the injected CO₂. MRCSP scientists also use constantly recording pressure and temperature gauges in numerous places in the system. Not only is the information collected from these gauges important to the research, it also helps ensure a safe and properly working injection system.

Once injection has ceased, the MRCSP team will conduct additional monitoring to measure reservoir conditions as they stabilize. The overall monitoring effort prior to, during and after injection is designed to ensure that we understand the behavior of the CO₂ and that the CO₂ is safely and permanently sequestered. Sampling of the cement and steel casing used in the well bores in addition to sampling of reservoir brines (the salty water indigenous to the formation) will provide data on the long-term changes to the well system and geologic formation that occurred due to injection of CO₂. Sensors dropped into the well bores on wires (wireline tools) including the use of crosswell seismic analysis will be carried out after injection to help determine the location of the injected CO₂ between the injection and monitoring wells.

Additional Value of the Extended Test

There are a number smaller scale tests being carried out across the US as part of the US Department of Energy’s seven Regional Partnerships including the MRCSP with the objective of evaluating various aspects of geologic sequestration and helping to ensure that it can be done safely and effectively. Larger scale tests are needed to further understand the behavior of the CO₂ when injected into deep geologic formations.

A key limitation to conducting larger scale tests in many areas of the country is that the only option for carbon dioxide is to purchase it through the commercial market. This is very expensive and it can be difficult to obtain large volumes of CO₂ for short time periods due to competing uses for food preparation, manufacturing and other commercial applications. The Michigan Basin test location has ready access to large amounts of pure carbon dioxide because of its proximity to the Turtle Lake natural gas processing plant operated by MRCSP partner DTE Energy and other natural gas processing plants in the area. In addition, the Michigan Basin test site is near compression and pipeline facilities owned and operated by Core Energy that support enhanced oil recovery operations in the area. Without this access to large amounts of CO₂ at reasonable cost and the compression and transport infrastructure already in place to support injection, it would be very costly to conduct a test of similar size.

Please look back here for periodic updates on this injection test.

The Midwest Carbon Sequestration Partnership completed an injection test of 10,240 tons of carbon dioxide (CO₂) earlier this year. At the time of completion, it was the largest deep saline reservoir test in the country.

The results from this test are promising: they show that the predictive model of the reservoir, which was developed prior to injection, was reasonably accurate and the test confirms the effectiveness of certain monitoring techniques. By design, this test involved only a small amount of carbon dioxide, but it was very useful in analyzing many aspects of the geologic sequestration concept. The observations from this test also help identify specific areas for further assessment in geologic storage tests.
This snapshot describes the methods that were used to assess the injection test and a preliminary review of the results.

Fall 2008

The Midwest Carbon Sequestration Partnership completed an injection test of 10,240 tons of carbon dioxide (CO₂) earlier this year. At the time of completion, it was the largest deep saline reservoir test in the country.

The results from this test are promising: they show that the predictive model of the reservoir, which was developed prior to injection, was reasonably accurate and the test confirms the effectiveness of certain monitoring techniques. By design, this test involved only a small amount of carbon dioxide, but it was very useful in analyzing many aspects of the geologic sequestration concept. The observations from this test also help identify specific areas for further assessment in geologic storage tests.

This snapshot describes the methods that were used to assess the injection test and a preliminary review of the results.

Basic Approach to Testing and Monitoring
The above graphic illustrates how the geologic understanding and confidence in the storage and containment system improved as various steps in the project are completed:
  1. The first version of the simulations involved the development of a computer model of the geologic reservoir, based on information that was already known about the region. The model was used to predict the behavior and movement of the carbon dioxide once it is injected. Reservoir simulations similar to those used by oil companies in oil and gas production were used with special features that account for geology and behavior of carbon dioxide in the ground.
  2. During the site characterization, the test well was drilled to collect site-specific data. Core samples and direct measurements from the well were taken and analyzed to determine porosity, permeability and other relevant information. (See the November 2006 Snapshot below for pictures of drilling and core sample collection.)  Note: Porosity is the amount of space between grains of rock and permeability is the measure of fluid flow potential in the rock based on connectedness of the pore spaces. A good storage reservoir has sufficiently high porosity and permeability, also known as injectivity. A good cap rock has low porosity and permeability and acts as a barrier to prevent carbon dioxide from rising to the surface. 
  3. The data collected from the test well were used to develop a site-specific model. The results of the model aided the design of the injection test and monitoring program, such as the appropriate location for monitoring wells (see below).
  4. After the injection test, the model will be calibrated based on actual field injection rates. The model predictions will be compared to field observations to validate and further refine the model.  
Placement of the Well and Monitoring Systems

As shown in the graphic below, a broad range of monitoring techniques were used to monitor for changes in the surface and subsurface in response to the injection test. The graphic also illustrates the geologic formations under the test site (Note that it is not drawn to scale). Enhanced oil recovery (EOR) operations take place at this site through wells located about 5,800 feet below the surface. A new well was drilled so that the carbon dioxide could be injected into a saline formation located approximately 3,500 feet below the surface for the test, and then be used later for EOR. Detailed reservoir modeling indicated that the injected carbon dioxide would not extend more than approximately 500 ft from the injection well and the pressure increase from injection would not extend more than 1,000 ft from the injection well. Two existing wells located 500 and 1,800 ft away from the injection well were reworked so that they could be used as monitoring wells. The placement of these two wells is indicated on the map below.
This figure shows the location of the injection well and the two monitoring wells. Monitoring well C3-30A was used for acoustic emissions (micro-seismic activity), crosswell seismic survey, fluid sampling and wireline logging. Monitoring well 2-30 was used for acoustic emissions. Well C3-30A was drilled in the direction of the injection well; thus, at the surface the well is ~850 ft away but only ~500 ft away at the bottom.

Monitoring Examples and Preliminary Results

Direct Measurements at the Well
The pictures above show the injection well in operation during the injection test. Pressure and temperature are continuously monitored at the surface and additional measurements are taken at the base of the well in the injection zone. Automatic sensors report any unusual change in pressure, temperature or volume.

10,240 metric tons of carbon dioxide were injected between February 18 and March 8, 2008 (including time taken for the mechanical integrity test required by the regulators). During this period, the injection rate was increased up to 600 metric tons/day. Bottomhole pressures were 2,000 to 2,020 pounds per square inch during the injection and remained generally stable. The data suggest that even higher injection rates could be sustained at this site. During injection, pressure was monitored at one of the monitoring wells located approximately 500 feet from the injection well. As shown in the figure below, the pressure declined within a few days after injection stopped.
Crosswell Seismic Survey

Crosswell seismic surveys are used to determine if the area of CO₂ spreading in the between the injection and monitoring wells can be observed. To conduct the test, a transmitter is temporarily placed in one well and sonic receivers (geophones) are placed in a monitoring well. Sound waves are bounced through the rock and a two-dimensional image is created based on the amount of time is takes for the wave to be received. An initial survey is conducted before any carbon dioxide is injected in order to create a baseline image of the rocks. Once the injection is finished, a second seismic survey is conducted to determine where the carbon dioxide has moved. The image below shows the placement of the equipment and the initial two-dimensional image of the rock.
Fluid Sampling

Samples of the groundwater, or formation fluid, were collected from the injection well and the nearest monitoring well before injection took place. The formation fluid in Michigan was very salty and had a total dissolved solids (TDS) concentration greater than 300,000 parts per million (the standard for potential drinking water supplies is 10,000 parts per million). About one month after injection was finished, new samples were collected over an 18-hour period and then analyzed. Brine samples collected from the monitoring well showed some subtle changes in calcium and magnesium concentrations. Calcium and magnesium are the primary elements in dolomite, the most abundant mineral in the injection interval. Although there was no direct observation of CO₂ reaching the monitoring well, these changes maybe indicative of the chemical processes that are likely to occur in the vicinity of the CO₂ injection zones. The picture below shows a researcher preparing fluid samples in the field for transport back to the lab.
Reservoir Modeling Results and Comparison to Measured Results

The following images show the output from the initial reservoir simulations for the project. The model predicted that during the injection test with about 10,000 metric tonnes of carbon dioxide, much of the injected carbon dioxide will not migrate more than about 500 feet from the injection well before stabilizing.
Pressure data collected at the injection and monitoring wells are consistent with the predicted pressures from the model. This suggests that the model was a reasonably good predictor of carbon dioxide behavior. In a typical project, the results from the first injection tests would be used to calibrate – or refine – the model and then the refined model would be validated based on new collected data.
Next Steps

Data collected during and after the injection test are being fully analyzed. Preliminary review shows that the model was an accurate predictor of the hydraulic response to injection. These data will be used to further refine and calibrate the model and also determine the applicability of various monitoring methods deployed at the site.

Spring 2008

The Midwest Carbon Sequestration Partnership has made considerable progress on the Michigan field test. A permit to inject carbon dioxide was issued in January 2008 and injection began on February 21, 2008. About 10,000 tons of carbon dioxide were injected as planned over about a six week period. The research team is now monitoring the injected carbon dioxide and analyzing the results.
This picture shows the well actively engaged in injection.
Before injection could start, the research team had to complete final integrity testing, known as the Mechanical Integrity Test or MIT to verify that the well was constructed correctly and that injection would not risk contaminating groundwater supplies. One of the tests involved developing a set of temperature logs across the well. The following picture shows the equipment set up for this test.
Once the injection test began, MRCSP invited members of the media to come to the injection site and learn about the test firsthand. The following pictures show the media being briefed on the project and then out in the field looking at the injection well and monitoring equipment.
Many of the same materials that were used in the earlier public meeting were used in the media briefing.

Several newspapers and a local radio station picked up the story. You can view the articles here.

July 2007

The Midwest Regional Carbon Sequestration Partnership (MRCSP) held a meeting for the public on July 18, 2007 to share information on a planned project to demonstrate the potential for deep geologic storage of carbon dioxide as a means to address global climate change. The proposed demonstration site is an existing oil and gas field, located off Sparr Rd. west of Sawyer Rd. in Charlton Township.

The meeting included a presentation on carbon sequestration and the Michigan Basin project (because of size, the presentation is included here in two parts: part 1 and part 2) and several stations that included poster displays and handouts. The materials that were used at the information stations included several fact sheets and posters already posted on this website and are not shown again here.  
  1. Presentation (because of size it is in two parts: part 1 and part 2)
  2. Chart on key steps in the process
  3. Site map
  4. Poster showing the purpose of the sequestration project
  5. Article published by the Bay City Times
  6. Poster of monitoring techniques
  7. Michigan geology panel
  8. Explanation of rocks and minerals
  9. Panel porosity and permeability (panel 1, panel 2)
  10. Graphic of rock formations
  11. Factsheet on Michigan Basin project
  12. Fact sheet on monitoring techniques
  13. A video entitled An Introduction to Carbon Capture and Sequestration
The meeting was open to all interested individuals, groups, and agencies. Two sessions were offered: from 2:00 to 4:00 p.m. and from 7:00 to 9:00 p.m. The meeting drew about 20 participants in the afternoon and 14 in the evening.

The following photos were taken at the meeting. The first shows one of the presentations and the second is of one of the experiments.

November 2006
In November, the MRCSP team worked with a crew to drill the test well for the research project. This slideshow explains the various stages of the process including drilling and core sampling.

  • Slide Title

    The Michigan Basin test will take place in Otsego County located, as indicated here, on the northern rim of the Michigan Basin.

  • Slide Title

    The field in which the test will take place. The test location is just at the edge of the woods at the end of the road in the foreground.

  • Slide Title

    The drilling rig is being brought in and assembled for use at the test location.

  • Slide Title

    As part of the field test, researchers took a core sample of the rock formations in which the well is located. This picture shows the rig crew assembling the special drill bit for the core sampling.

  • Slide Title

    The special drill bit for core sampling has a hollow center. As the drill moves through rock it cuts a cylindrical sample that shows the various layers and composition of rock at precisely measured depths. In this picture, the drilling crew is adding extensions to the drill bit as it moves deeper into the ground. As the core is cut from the rock it is lifted out of the well in a case to help keep it intact. The picture below shows a core sample packed in protective casing.

  • Slide Title

    This is a close-up shot showing about two and a half feet of core sample that was taken at a depth pf 3,085 feet. The picture below shows the twenty or so feet of core immediately above located at the depth of 3,060 feet and going to 3,084 feet. The core is marked when it comes out of the well so that it can be reassembled in the lab. These rocks are from the Amherstburg formation which is a dense limestone with many fossil fragments embedded in it. This rock is considered a confining layer or cap rock because it has very little pore space.

  • Slide Title

    Going 400 feet deeper than the samples above, the crew drew a core sample from the Bass Islands rock formation. The core pictured here was taken from the interval located between 3,460 to 3,482 feet underground. Here the Bass Island formation consists of a type of porous rock known as a brown dolomite.

  • Slide Title

    Now that the core sample is in the lab, researchers with the MRCSP will run tests to determine the porosity and injectivity of different layers as well as the suitability of other layers as caprock.

Share by: