JPT | 14 August 2017
Risk Assessment of Fluid Migration Into Freshwater Aquifers in Colorado Basins
Wellbore-construction methods, especially casing-and-cementing practices for the protection of freshwater aquifers, have been reviewed in the Piceance, Raton, and San Juan Basins in Colorado. The assessment confirms that natural-gas migration occurs infrequently but can happen from poorly constructed wellbores. Analysis confirmed no occurrence of hydraulic-fracturing-fluid contamination. The significance of these results is to help quantify the risks associated with natural-gas development as related to the contamination of surface aquifers.
The prevention of contamination of freshwater aquifers has been a prime concern in drilling operations since the inception of drilling. Surface casing has long been the primary barrier to prevent contamination of freshwater aquifers through wellbores. The probability of leakage into aquifers from wellbores during shale development has a wide range of estimates, complicated by the presence of hydrocarbons at shallow depths in many parts of the world. An earlier paper reviewed the process and outcomes of a study for the Wattenberg Field in the Denver-Julesberg Basin. This study continues the examination of the contamination of aquifers in the subsurface during the completion and the production phases of the well and quantifies the risk of contamination of aquifers through failure of the wellbore for three other major basins in Colorado, the Piceance, Raton, and San Juan Basins. This synopsis focuses on the assessment of the Piceance Basin.
Barrier Definition. Common vertical, deviated, and horizontal subsurface wellbore-barrier designs were grouped and ranked on the basis of the risk of multiple barrier failures (Fig. 1). For the sake of clarity, pressure monitoring of the casing annulus [surface annulus pressure (SAP)] was not assumed to be an additional barrier during the production phase even though it is frequent and often required by state regulations.
Fig. 1—Wellbore-barrier categories ranked from highest risk to lowest risk.
Well-barrier designs can vary from field to field depending on geology, trajectory, depths, anticipated pressures, expected hydraulic-treatment rates, and estimated production rates. Whether a well is horizontal, vertical, or deviated has no significance with respect to the ultimate protection of freshwater aquifers because the wells are designed to protect the shallow vertical section of each oil and gas well. Multiple barriers must be in place near the depth of the freshwater aquifer to prevent breaching of a single barrier potentially leading to contamination.
Failure Definition. This study defines two types of barrier failure—potential barrier failures and catastrophic barrier failures.
Potential barrier failures are the breakdown of a single or multiple barriers in a wellbore that did not result in the contamination of freshwater aquifers or surface soil from hydrocarbon or fracturing-fluid migration but required remediation of the failed barrier to further enhance the nested barrier system of the well.
Catastrophic barrier failures are the breakdown of a combination of various wellbore barriers (casing, cement, and hydrostatic pressure of annular fluids) protecting freshwater aquifers during hydraulic fracturing or production phases resulting in the contamination of freshwater aquifers or surface soil.
Risk Assessment of Oil and Gas Wells in the Piceance Basin
The Piceance Basin underlies western Colorado. Garfield County comprises the core of oil and gas exploration for the Piceance Basin in Colorado and is the focus of this study.
The first exploratory drilling operations were started in 1935; however, concentrated oil and gas exploration of the field did not begin until 2000. Wells are drilled to depths of 6,000–8,000 ft subsurface.
Horizontal wells began around 2008, but horizontal-drilling activity is negligible compared with the vertical- and deviated-well counts because of the complexity of drilling a horizontal well at increased depths of 10,000- to 12,500-ft true vertical depth. Wells in the Piceance Basin are subject to higher corrosion rates because of elevated levels of total dissolved solids (TDS) in the produced water and the presence of corrosive gas. In addition, cementing wells can be problematical because of fracturing in the Wasatch Group, which is above the Williams Fork Formation target zone. The Wasatch Group is composed of interbedded shale and sandstone and is highly fractured because of structural alterations and thermogenic gas migration from the underlying Williams Fork Formation.
Piceance Basin Water Sourcing. A defined geologic boundary between freshwater aquifers and deeper hydrocarbon formations is not present in Garfield County. Water is sourced from surface water, unconsolidated alluvial aquifers that are shallower than 60 ft subsurface, and deeper water wells that source fresh water from bedrock in the Wasatch Group at maximum depths of 600 ft.
The Wasatch Group has evidence of natural fractures that can act as a conduit to shallower depths from deeper and more-mature hydrocarbon deposits. The deepest water wells in Garfield County are drilled to 600 ft, sourcing water from the Wasatch Group. Water wells that are drilled into the bedrock of the Wasatch Group have the potential to test positive for thermogenic gas without any offset oil and gas wells contributing to thermogenic gas migration to the aquifer. It is challenging to ascertain the origin of thermogenic gas that appears in water wells, because of the complexity of the underlying strata in Garfield County.
Piceance Basin Data Sourcing and Assumptions. Oil- and gas-well data were collected for 10,998 wells completed between 1935 and mid-2014 in Garfield County. Shallow surface-casing-setting depth was defined as a depth less than the deepest water well in the field at 600 ft.
Potential barrier failures were defined as any cement remediation performed on the production casing, intermediate casing, or surface casing or as presence of SAP.
Catastrophic barrier failures were defined as wells that had barrier failures that directly caused a conduit for hydrocarbon migration to freshwater aquifers of the upper Wasatch Group or to alluvial aquifers at shallow depths, which was corroborated by isotopic and compositional analysis from an offset water well.
Piceance Basin Potential and Catastrophic Barrier Failures. All wells were categorized on the basis of their original casing and cement. Potential barrier failures were identified by any cement remediation of any casing string or by evidence of SAP. Potential barrier failures were identified in 377 of 10,842 originally producing wells in Garfield County. Category 8 wells had the highest potential-barrier-failure rate of 30.00%, occurring in 18 of 60 wells. Even though this design has deep surface casing and an intermediate-casing string, the top of the production-casing cement was not above the top of gas. Higher-risk Category 2 wells had an 8.33% potential-barrier-failure rate, occurring on four of 48 wells, followed by Category 5 wells, which had a 6.99% potential-barrier-failure rate, occurring in 125 of 1,789 wells. This design has deep surface casing, but the top of the production-casing cement was not above the top of gas.
Categories 6 and 7 (lower risk) had lower potential-barrier-failure rates of 2.33 and 3.01%, respectively. Even though these wells had production-casing-cement tops above the top of gas, they demonstrate the challenging geologic conditions that are present in the Wasatch Group, confirming the difficulty in creating effective production-casing-cement isolation and indicating challenges in preventing SAP from shallow hydrocarbon deposits.
Higher concentrations of potential barrier failures occurred near the core of oil and gas exploration. Of wells that were originally completed in 2003, 18% had potential barrier failures, which represents the most potential barrier failures for wells completed in a calendar year.
The higher potential-barrier-failure rates experienced for lower-risk wellbore-barrier designs in the field are because of corrosion or ineffective cement coverage behind the casing strings. High carbon dioxide (CO2) mole fraction and higher relative TDS from the produced water can lead to corrosion of the carbon-steel pipe wall if untreated. Proper treatment of this corrosion potential is needed to prevent casing leaks and deterioration of the pipe wall.
Nine of 10,842 originally producing wells were identified as having catastrophic barrier failures related to hydrocarbon migration to freshwater aquifers in the Piceance Basin. All nine wells had experienced high SAP before thermogenic gas detection in offset water wells. No evidence of hydraulic-fracturing-fluid migration to freshwater aquifers or surface soil was found.
Catastrophic barrier failures occurred in two Category 3 wells, two Category 5 wells, four Category 6 wells, and one Category 7 well. Seven catastrophic barrier failures occurred on wells that had the top of production-casing cement above the top of gas in the basin. This demonstrates the challenges in effectively isolating shallow gas shows in the Wasatch Group and the challenges of higher carbon-steel corrosion rates and of annular-hydrostatic-pressure barriers in the field.
Catastrophic-barrier-failure rates were observed to be common in moderate- to low-risk wellbore-barrier designs because of the challenges of combating corrosion of the production casing, of effectively isolating hydrocarbon migration with cement in the casing annulus, and of creating effective annular hydrostatic barriers at shallow depths. No evidence of fracturing-fluid migration to freshwater aquifers was detected in the study. Formations that are hydraulically stimulated are at depths greater than 4,000 ft subsurface. Even with the highly fractured nature of the Wasatch Group above the top of the Williams Fork Formation, it still acts as a solid geologic barrier, preventing vertical growth of artificially stimulated fractures to freshwater strata.
Piceance Basin Existing Conditions. Six percent, or 602, of the 10,507 existing producing or shut-in wells in Garfield County currently have shallow surface-casing-setting depths in relation to the deepest water well drilled in the county. Of these 602 wells, 143 currently have higher-risk Category 2 and 3 well-barrier designs. These designs do not have the top of production-casing cement above the surface-casing shoe.
Of wells in the sample, 3.48% experienced SAP, had cement remediation, or had a combination of both. This higher potential-failure rate relative to the Wattenberg Field is explained by the shallow gas shows from the Wasatch Group, the difficulty in eliminating shallow gas shows with effective cement coverage, and higher rates of corrosion of the production casing because of the relatively higher TDS from the produced water and higher mole percent CO2 in the produced gas. In order to isolate annular migration, it is recommend-ed that operators extend production-casing cement above the preceding casing shoe and routinely perform chemical batch treatment of wells to reduce the effects of corrosion of the pipe walls.
This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 181680, “A Continued Assessment of the Risk of Migration of Hydrocarbons or Fracturing Fluids Into Freshwater Aquifers in the Piceance, Raton, and San Juan Basins of Colorado,” by C.H. Stone, SPE, A.W. Eustes, SPE, and W.W. Fleckenstein, SPE, Colorado School of Mines, prepared for the 2016 SPE Annual Technical Conference and Exhibition, Dubai, 26–28 September. The paper has not been peer reviewed.