2.5.2 Possible pathways and common causes for CO2 leakage in an abandoned wellbore

In an abandoned well, there are various pathways for the leakage of CO2 to overlying permeable intervals or to the surface (Nygaard and Lavoie, 2010; Duguid and Tombari, 2007; Syed and Cutler, 2010). These include:

  • CO2 migration through the pathways and/or pores of the well cement (both for cement plugs in the wellbore and for the cement behind the casing, in the production casing/outer casing annuli),
  • CO2 leakage through the annular space between the production casing and the cement behind the casing,
  • CO2 leakage through the annular space between the cement behind the casing and the formation,
  • CO2 leakage through damaged casing and
  • CO2 leakage through damaged bridge plugs or other mechanical plugging equipment set in the wellbore.

The most common cause of leakage from abandoned wells is failure of the cement to prevent CO2 flow. This can be either due to failures in the proper application of primary casing cement jobs, cement squeeze jobs and cement plugs or due to cement deterioration when the cement is exposed to both CO2 and the water acidified by the presence of CO2. Although a study on the analysis of the cement in an oil well with 30 years of CO2 exposure showed that the cement retained its capacity to prevent any significant flow of CO2 (Carey et al., 2007), various laboratory studies (Carey et al., 2010; Brandvoll et al., 2009) demonstrate the deteriorative effect of CO2 and CO2-brine mixtures on cements. Research has demonstrated that there may be CO2 leakage even from wells which are properly cemented and abandoned, mainly because of the weak acid formed by the interaction of CO2 with formation water which deteriorates the cement quality.

A study performed to investigate the factors affecting wellbore leakage in Alberta, Canada (Watson and Bachu, 2009), indicates that the factors which have the greatest impact are:

  • Geographical location: The study showed that wells at certain regions demonstrated a higher probability of leakage. This may be due to the geological setting, site-specific problems with drilling, completion and cementing, and varying operating practices applied by different companies operating in different regions.
  • Wellbore deviation: Compared to vertical wells, the occurrence of leakage was higher in deviated wells. The reason may be improper casing cementation in deviated wells due to casing centralisation problems and cement slumping, increasing the probability CO2 migration behind the casing.
  • Well type: Cased wells have higher potential for CO2 leakage compared to drilled and abandoned wells without casings. The reason can be more stringent abandonment procedures required for drilled and abandoned wells and also, additional pathways for CO2 migration existing in cased and perforated wells.
  • Abandonment method: For cased and completed wells, the most commonly used abandonment technique in Alberta is setting bridge plugs capped with cement. The study showed that 10% of these types of abandonment applications will fail in long term (hundreds of years). In comparison, other methods such as setting a cement retainer and squeezing cement through perforations or placing a cement plug across perforations is expected to have a lower failure rate. The damaging effect of CO2 on the elastomers and metal elements of bridge plugs may also add to the failures of such mechanical plugging equipment.
  • Oil price and regulatory changes: As the oil price increases, the drive to achieve more with lower cost may have led to lower quality of performed primary cement jobs. Regulatory changes, imposing stricter procedures for well abandonment, are expected to improve the abandoned well integrity.
  • Uncemented casing/hole annulus: Low cement tops behind casings were found to be the most important factor for CO2 leakage in the Alberta study. The absence of cement behind the casing increased the external corrosion of the casing significantly and also allowed the CO2 to move up the uncemented pathway.