A salt solution well’s design involves the use of two or more steel pipe columns (casing strings) nested within each other. To initiate the solution process within a expanding salt cavern, fresh water or undersaturated brine (known as the feed solvent) is pumped into the well, creating a void within the targeted salt formation. The resulting brine is then extracted to the surface for processing and recovery.
The initial step in salt solution mining is to drill a borehole with a diameter large enough to accommodate all the required pipes (casing). As a result, solution wells are wider than most oil, gas, or water wells. Near the surface, the borehole widens further to facilitate the installation of multiple concentric layers of pipe casing. The outermost layer, referred to as the surface casing, is cemented in place to prevent any leakage and contamination of nearby groundwater. Typically, the surface casing does not extend all the way down to the cavern roof. The final casing string, also cemented in place, is set at a depth below the top of the target salt to ensure a sufficient salt roof remains intact as the cavity expands. In modern wells, there is an additional inner casing string that controls the thickness of the fluid blanket.
Usually, one or more non-cemented casing strings, known as tubing strings, are inserted inside the final casing string. In modern cavern designs, these non-cemented tubing strings often extend to a depth near the planned base of the cavern. This design allows for adjustment as the cavern expands and collects insoluble debris, which can accumulate at the cavern’s bottom, sometimes reaching depths exceeding 20 meters, depending on impurity levels (dolomite-anhydrite residues and fragmented roof blocks).
When subjecting salt to undersaturated crossflow within an expanding solution cavity, vertical leaching occurs 1.5 to 2 times faster than horizontal leaching. This is due to the lower density of fresh water compared to brine, causing fresh water to accumulate in the upper part of the cavity. Vertical leaching can be even faster if gas inclusions are present in the cavity water, increasing buoyancy. To prevent rapid upward dissolution and the eventual collapse of the cavern roof and its cemented casing, an inert low-density fluid blanket is pumped into the solution cavern. Roof blanketing, sonar imaging, and cavern shaping are typical practices in contemporary commercial solution mining operations.
The blanketing fluid is inert, ensuring it does not dissolve the target salt. It is always less dense than the feed solvent or the produced brine, floating to the cavern’s top to insulate the roof from dissolution effects. Older blankets often consisted of various types of oil, such as diesel or crude oil, but modern ones increasingly use other fluids like LPG, nitrogen, and compressed air. Nitrogen and, less frequently, compressed air are preferred due to their lack of environmental and safety concerns associated with hydrocarbon blankets. Creating a blanket involves adding a dedicated column of pipes in the well string, and the blanket is typically pumped in and out through the annular space between the outer and middle casing pipes.
(a) In the direct brine circulation method, the feed solvent is introduced through the tubing string, while brine is extracted through the space between the tubing string and the outer casing. This results in a cavern shape that tends to be cylindrical with a slightly expanded lower section.
(b) In the reverse-circulation approach, the feed solvent enters the cavity through the annular space, and brine is withdrawn through the tubing string. This typically leads to a cavern with a wider top than its base, often referred to as a ‘morning glory’ shape. In both cases, the outermost casing allows the introduction of the blanket solution into the uppermost part of the cavern, thereby safeguarding the cavern’s roof. Adjusting the volume of the blanket solution within the cavern can be utilized to influence the cavity’s shape.
There are several methods for developing and shaping caverns. In the direct circulation method, feed solvent is injected through the tubing string, while brine is withdrawn through the annular space between the tubing string and the final casing. In the reverse-circulation method, feed solvent enters through the annulus, and brine is withdrawn through the tubing string. Reverse circulation tends to create a “morning glory” type cavity with a wider top than base, especially in cavities formed without the aid of a fluid blanket. In contrast, direct circulation typically results in a more cylindrical cavity. During cavern excavation, the volume of the blanket fluid can be adjusted to shape the cavity as desired.
In-depth discussions of the engineering aspects of well design and cavern shaping are beyond the scope of this geological focus. Various authors, such as Remson et al. (1966), have outlined general procedures for controlling cavern shape, and numerous articles by Shock (1985) and Jacoby (1974), as well as bibliographies like those by Dahl (1985) and Nigbor (1982), cover basic solution mining techniques for different salts. Combining various methods and advanced solution mining approaches are now well-understood practices for achieving desired cavern geometries and well-to-well interconnections (API 1994).
If the objective of a solution mining operation is to obtain a specific brine product rather than the cavern itself, the extracted brine is usually evaporated or treated at the surface to produce crystallized products. In such cases, maintaining an “almost saturated” brine product in the rising casing string is crucial. Dilute brine is costly to process, while supersaturated brine can lead to unwanted precipitates clogging the raising string.
Initially, when drilling a single-hole solution mining well, the flow rate through the cavity and up the pipe is typically kept low to maintain higher brine concentration levels. As the cavity’s surface area and volume increase over time, the extracted brine’s flow rate can be gradually increased. In approximately a year, considerably higher flow rates can be sustained. Successful brine wells can have long lifespans, lasting up to 30–40 years. However, if the solution process is continuously maintained in a single well, the area around the well may eventually subside, or the borehole roof supporting the casing strings may weaken and collapse, disrupting brine flow. Even if the roof does not cave, salt can dissolve around the casing string, leading to corrosion and eventual collapse into the cavity.
Historically, in Western countries until the 1950s, and much later in the former Soviet Union, captive wells designed for brine recovery, rather than storage, were often expected to be abandoned or to fail. Many such wells initially performed well but eventually failed due to casing string breakage, corrosion, roof caving, or subsidence-related damage (Deutsch 1978). Failures could occur shortly after a well became a high-capacity producer or decades later. Operators of brinefields discovered that some of their oldest wells were the most reliable. Many captive brinewells experienced difficulties during their early stages, but if they survived the “childhood diseases” of roof collapse and casing shear, the likelihood of failure decreased as they aged. Some older brinefield wells in bedded salt formations, which had been in production for 20–30 years, appeared to be resistant to caving and were only abandoned when the target salt layer was depleted. Predicting which wells would endure in a brinefield was challenging. An analysis of around 60 wells drilled between 1887 and 1925 clearly showed the trend of long-lived survivors. One possible explanation for the longevity of such wells is that early cave-ins distributed fallen rock in a way that protected the lower part of the tubing string from damage by subsequent cave-ins. Alternatively, as the roof of a growing salt cavity spanned a greater distance, it could either collapse with detrimental effects on the well or gradually subside until it was supported by silt, projections, and debris at the bottom of the cavity. In the latter scenario, a brine well could remain a good producer even with a shallow cavity. Well failure or poor performance exhibited some randomness and could occur unexpectedly. As a result, and due to slow initial production rates, most operators of captive brinefields employed numerous wells, abandoning non-producers and bringing new ones online.
In all cases, mud and other insoluble materials, as well as some salt debris, accumulated as muddy breccias at the base of the expanding cavity. In cases where salt layers other than halite were the target and were enclosed within thick halite formations, the initial borehole was typically drilled through the target beds to a certain depth into the underlying halite mass. As the cavity expanded, it formed a sump beneath the target bed where insolubles could collect. In any solution cavity, it was necessary to raise the bottom casing string (a process known as snubbing) above the level of the insolubles; otherwise, the well would need to be plugged and abandoned. This was also true if the casing had been severely bent or damaged by rockfalls from the roof. These problems usually developed gradually, and the total mining expenses, including initial capital and well operation costs, were relatively low compared to conventional room and pillar extraction.
Many land areas situated above brinefields, which were active during the eighteenth or nineteenth centuries and often over-exploited, are susceptible to collapse today. However, few records exist regarding well placement or the brinefield’s exact location in older extraction areas (as seen in Hutchinson, Kansas). This lack of documentation leads to various environmental and civil engineering challenges in the vicinity of many former brinefields. Nevertheless, not all engineering and subsidence problems in near-surface salt areas are human-made. Salt naturally dissolves, and large collapse cavities form over time, often within days or weeks. Outside of areas above poorly managed or historical brinefields, distinguishing between anthropogenic sinkholes and natural dissolution cratering is exceedingly difficult. Ultimately, interpretations of causative factors often depend on the interests funding the consultancy.
