Principal Investigator

At a Glance

Modeling of CO2  injection in the Marcellus shale formation reveals the need for a very large number of injection wells—approximately 100 additional wells brought online every year—to store CO2  emissions from Pennsylvania’s five largest coal-fired power plants over a 40 year lifetime.


Research Highlight

Research in the Celia group has continued with modeling general aspects of carbon dioxide (CO2) injection1-7 into conventional formations and measuring methane leakage along old wells8-10. In addition, a study of the feasibility of injection of CO2 into depleted shale-gas formations has continued. The idea to inject CO2 into depleted shale formations has been advanced recently as an alternative or complement to injection into conventional reservoirs. Initial estimates from the literature indicated very large (static) storage capacities11,12; the current research focuses on realistic estimates of injection rates and the logistical feasibility of different injection scenarios. To address this issue, the Celia group conducted a thorough review of gas production data for the Barnett formation in Texas as well as two different regions of the Marcellus shale formation in Pennsylvania (Southwest Pennsylvania (SW PA) and Northeast Pennsylvania (NE PA)). The Celia group developed a model for multi-component gas flow within the various formations, including the movement of methane (CH4 ) and CO2. The model includes equations of state, sorption of both CH4 and CO2, and formation parameters derived from matching production data.

The results in Figure 2.1.1 show cumulative injected mass of CO2 into a typical well in the three formations as a function of time. The NE PA region has the best performance, because the formation is much thicker in the northeast than in the southwest, and because length of horizontal wells has become noticeably longer in NE PA. While the northeast location has the best behavior, the amount that can be injected into one well is still orders-of-magnitude smaller than the emissions from a typical coal-fired power plant: ultimate cumulative mass injected in one well is about 0.5 Million Tonnes (Mt) CO2, while the output from one large coal-fired power plant over that time (40 years) is on the order of 200 Mt CO2. It is worth noting the total mass injected has an upper limit due to the assumptions of the one-dimensional model; more detailed simulations may show some additional capacity, but the additional amount is expected to be insignificant.

Figure 2.1.1. Cumulative mass of CO2 injected as a function of time for a typical well in the Barnett, Southwest PA Marcellus, and Northeast PA Marcellus formations. Note that half of the ultimate capacity is reached after about 5 years for all three curves. (Figure courtesy of the Celia group.)

Large stationary sources in Pennsylvania were identified and the CO2 emissions were mapped to existing (both drilled and permitted) shale-gas wells. Figure 2.1.2 shows both sources and wells. If the CO2 produced by the large sources in southwest Pennsylvania is injected into wells in southwest PA, essentially all wells in the region (approximately 6,400) are needed over a time period of 40 years (more than 150 new wells each year). If the higher-injectivity wells in Northeast Pennsylvania are used instead, then the number of wells will be 3,800 (about 100 wells per year). Note that use of wells in Northeast PA requires longer pipelines to be built.

For either of these cases, it is reasonable to ask if it is more economically and logistically feasible to build a large pipeline west of the Illinois Basin, and use conventional reservoirs such as the Mt Simon formation to sequester the CO2. A comprehensive economic and feasibility study is currently ongoing, to clarify advantages and disadvantages among these different options for CO2 transport and injection

Figure 2.1.2. Stationary sources of CO2 (triangles with size scaled to amount of emissions), and existing and permitted shale-gas wells (all dots). The gray shaded region indicates the Marcellus shale formation in Pennsylvania. Purple dots (6,400 or them) identify wells needed to store 40 years of emissions from the five large sources in southwest Pennsylvania (orange triangles); green dots (3,800 of them) identify wells in northeast Pennsylvania to store the same emissions. (Figure courtesy of the Celia group.)



  1. Bandilla, K.W., M.A. Celia, and E. Leister, 2014. Impact of Model Complexity on CO2 Plume Modeling at Sleipner. Energy Procedia, 63: 34053415.doi:10.1016/j.egypro.
  2. Bandilla, K., M.A. Celia, J.T. Birkholzer, A. Cihan, and E.C. Leister, 2015. Multi-phase Modeling of Geologic Carbon Sequestration in Saline Aquifers. Ground Water, February 6, 2015 (online).doi:10.1111/gwat.12315.
  3. Zheng, Z., B. Guo, I. Christov, M. Celia, and H. Stone, 2015. Flow Regimes for Fluid Injection into a Confined Porous Medium. J. Fluid Mech., 767: 881-909. doi:10.1017/jfm.2015.68.
  4. Huang, X., K.W. Bandilla, M.A. Celia, and S. Bachu, 2014. Basin-scale Modeling of CO2 Storage using Models of Varying Complexity. Int. J. Greenh. Gas Control, 20: 73-86. doi:10.1016/j.ijggc.2013.11.004
  5. Guo, B., K.W. Bandilla, F. Doster, E. Keilegavlen, and M.A. Celia, 2014. A Verticallyintegrated Model with Vertical Dynamics for CO2 Storage. Water Resour. Res., 50(8): 6269-6284. doi:10.1002/2013WR015215.
  6. Guo, B., K.W. Bandilla, E. Keilegavlen, F. Doster, and M.A. Celia, 2014. Application of Vertically Integrated Models with Subscale Vertical Dynamics to Field Sites for CO2 Sequestration. Energy Procedia, 63: 3523-3531. doi:10.1016/j.egypro.2014.11.381.
  7. Kang, M., J.M. Nordbotten, F. Doster, and M.A. Celia, 2014. Analytical Solutions for Two-phase Subsurface Flow to a Leaky Fault considering Vertical Flow Effects and Fault Properties. Water Resour. Res., 50(4): 3536-3552. doi:10.1002/2013WR014628.
  8. Kang, M., C. Kanno, M. Reid, X. Zhang, D.L. Mauzerall, M.A. Celia, Y. Chen, and T.C. Onstott, 2014. Direct Measurements of Methane Emissions from Abandoned Oil and Gas Wells in Pennsylvania. Proc. Natl. Acad. Sci., December 8, 2014 (online). doi:10.1073/pnas.1408315111.
  9. Kang, M., E. Baik, A.R. Miller, K.W. Bandilla, and M.A. Celia, 2015. Effective Permeabilities of Abandoned Oil and Gas Wells: Analysis of Data from Pennsylvania. Environ. Sci. Technol., March 13, 2015. doi:10.1021/acs.est.5b00132.
  10. Kang, M., D.L. Mauzerall, D.Z. Ma, and M.A. Celia, 2015. Estimating and Mitigation Methane Emissions from Abandoned Oil and Gas Wells in Pennsylvania. Environ. Sci. Technol., under review
  11. Tao, Z. and A. Clarens, 2013. Estimating the Carbon Sequestration Capacity of Shale Formations using Methane Production Rates. Environ. Sci. Technol., 47(19):11318-11325. doi:10.1021/es401221j.
  12. Godec, M., G. Koperna, R. Petrusak, and A. Oudinot, 2013. Assessment of Factors Influencing CO2 Storage Capacity and Injectivity in Eastern U.S. Gas Shales. Energy Procedia, 37: 6644-6655.doi:10.1016/j.egypro.2013.06.597.