Colorado Mountain College, Spring Creek, Garfield County
In early February of 2003 a 24-foot wide sinkhole spontaneously opened on a soccer field at the Colorado Mountain College Roaring Fork Campus near Spring Valley, about 7 ½ miles southeast of Glenwood Springs. After filling by the CMC physical plant maintenance staff, the sinkhole reopened the next year and enlarged to about 35 feet in diameter.
The site is located near the southern end of Spring Valley (Figure 1). This interesting subsidence feature lies within the Carbondale Collapse Center. Spring Valley, originally a lake, has been interpreted as a drag feature that opened when a large block of basalt-capped redbeds (Maroon Formation) moved laterally, rafting on flowing salt that moved from high-pressure environments under Little Grand Mesa to lower pressure terrain at the axis of the Cattle Creek anticline at the floor of the Roaring Fork River valley. At the southern end of Spring Valley, evaporite dissolution and subsidence has created a complex and chaotic mixing of broken and intact basalt cap rocks, variable bedrock, and depression in-fill of sediments that have been mapped as “collapse debris” in the geologic maps completed by CGS in the late 1990s.
Figure 1. West view of Colorado Mountain College 2003 sinkhole showing proximity to campus buildings. Sunlight Ski Area in left background. Roaring Fork River valley is in left middle background below ski area and in front of first row of ridges.
On campus, rolling hills were graded flat to create the soccer field. While basalt blocks occur on the hill top above the field, a large spine of gypsum is exposed in the 25- to 30-foot cut slope that was required to make the field flat. An investigative drill boring at the sinkhole revealed gypsum of the Eagle Valley Evaporite at a depth of 65 feet, buried by sediments washed into a natural depression and 30 feet of fairly uniform, mostly reworked windblown sediments (loess) that can be seen in the walls of the sinkhole (Figure 2). The subsurface void chimneyed up through these sediments from a cavern in the evaporite bedrock below. Just a few months earlier in the summer and fall of 2002, this location was repeatedly driven over by a tractor pulling overlapping gang mowers to cut the grass of the sports fields.
Figure 2. South view of close up of sinkhole wall showing uniform silt sediments derived from windblown deposits (loess). CMC maintenance staff has started to fill sinkhole with broken road-pavement waste.
Iron Bridge (Rose Ranch) Golf Course at Roaring Fork Valley, Garfield County
On Sunday, January 9, 2005 a large sinkhole opened off of County Road 109, across the Roaring Fork River from Highway 82, between Glenwood Springs and Carbondale. The sinkhole occurred in the Iron Bridge community development, previously known as the Rose Ranch. In 2002, the CGS published a map of evaporite karst hazards for this area of the state. This site is located right of center in the block diagram shown in an existing evaporite karst hazards map by the CGS, Map Series 34, Collapsible Soils and Evaporite Karst Hazards Map of the Roaring Fork River Corridor. The underlying bedrock is Eagle Valley Evaporite and on this same property, about 1 mile south, an oil well (Shannon Oil Co. Rose no. 1) was drilled that encountered 60 feet of gravel, and 2,065 feet of gypsum, anhydrite, and siltstone before drilling through 935 feet of halite (salt) where the hole was stopped.
The sinkhole opened up at the clubhouse golf cart maintenance and storage facility (Figure 1). Reportedly, a small hole, about 10’ by 10’, opened very early Sunday morning that quickly enlarge to a 42-foot diameter and 40-foot deep sinkhole by the middle of the day. The opening undermined the southwest corner of a vinyl-tent Quonset where electric golf carts were being stored for the season. Apparently at least two of the small vehicles were lost into the sinkhole. Exposed in the hole were a ruptured 24-inch black plastic corrugated storm-sewer pipe and two smaller PVC pipes of unknown utility.
Figure 1. Ironbridge golf course sinkhole in 2005. View is down the Roaring Fork River valley towards Glenwood Springs. Redbed cliffy valleyside is the Maroon Formation. The valley bottom is the Eagle Valley Evaporite. Photo by Jon White.
There was some discussion that a possible leak in the storm sewer could have caused this sinkhole. While limited leakage may have reduced the strength of the void ceiling, being a sole cause of the sinkhole is unlikely. The soils exposed on the sinkhole wall are alluvial fan deposits and may be hydrocompactive, a type of collapsible soil (link to collapsible soil), but there was no report of earlier settlement and no evidence of distress related to earlier ground settlement surrounding the sinkhole location. The void had propagated upward into the alluvial fan soils in the recent geologic past from the chimneying of material down into a large dissolution void in the evaporite bedrock. The void size had to exceed the volume equivalence of the sinkhole opening - 40 by 40 by 40 feet). The void was of such size that, when the roof collapsed, the displacement of air cause a pneumatic “geyser” effect, and a small cloud of finer soil particle (coarse sand and dust) erupted and covered the ground around the sinkhole edge. It is likely that the overall grading and trenching for the storm sewer at this location weakened the ceiling of the existing subsurface void and accelerated the cavern’s propagation to the surface.
An interesting side note in this site visit. At the time of the CGS inspection, representatives of a local contractor that did most of the excavation for this project arrived to view the sinkhole. In our discussion they mentioned that they, many times, encounter voids in the soils in the area of the Roaring Fork Corridor during grading and trench excavations (Figure 2). Most are quickly backfilled and forgotten, complying with the wishes of the clients to keep it “quiet.” At times, this has occurred even under foundation footprints.
Figure 2. Utility construction in the Roaring Fork River valley uncovered this large void in river gravel that propagated upwards from a dissolution void in the underlying Eagle Valley Evaporite. Photo courtesy of Steve Pawlak, HP Geotech.
Lykins Formation strike valley along the Front Range, Larimer County
The Permo-Triassic Lykins Formation redbeds can contain massive gypsum deposits, up to 50 feet thick. Dissolution of those beds and some of the thin algal limestone is the unit is responsible for many sinkholes and ground subsidence features inside the main Dakota Sandstone hogback that marks the boundary of the Eastern Plains and the Front Range.
On May 1, 1999 a sinkhole spontaneously opened on the westbound shoulder of U.S. Highway 34 within the Colorado Department of Transportation (CDOT) right-of-way, about 3 miles west of Loveland. Just north of the sinkhole the beds of gypsum were historically mined by the United States Gypsum company. The mine has closed and the property is being redeveloped as a gated community. The roadside sinkhole was approximately 25 feet in diameter and 20 feet deep. Eye-witness reports that the material simply disappeared down a hole that angled slightly to the north. CDOT drilling revealed large voids within the Forelle Limestone member of the Lykins formation. The Forelle can be in an advanced state of karstic dissolution and brecciation because of the proximity of more soluble gypsum beds below.
Topographic lowering and closed depression are common within the Lykins valley along the Front Range. Where streams exit the foothills, the adjacent narrow water gaps in the tilted hard sandstone of the Dakota hogback are favorable location for dams. Two prominent reservoirs have been filled on the Lykins: Carter Reservoir and Horse Tooth Reservoir. The U.S. Bureau of Reclamation has had to mitigate seepage problems at both, related to gypsum dissolution. Sinkholes open frequently at the south end of Horse Tooth during seasonal drawdown of lake levels.
The thickest and most exposed gypsum in the Lykins Formation is exposed in the Table Mountain area in Larimer County near the border with Wyoming. Thick gypsum beds are exposed at the surface (Figure 1) and there are several historic gypsum mines in the area. Karst features also exist. There are several sinkholes (Figure 2) and a small creek is captured into a gypsum cave in this same region. This area will soon be open to the public as Larimer County’s Red Mountain Open Space.
Figure 1. Upper and lower beds of gypsum exposed on the more gently east-dipping limb of the assymetrical Sand Creek Anticline. Table Mountain is shown in background. Sinkholes occur within the small valley in center of photo between the chalky-white gypsum beds. Photo was taken from the top of the upper gypsum bed with the view to the south. Photo by Jon White.
Figure 2. One of many sinkholes in the Larimer County Red Mountain Open Space where the Lykins Formation contains thick beds of gypsum. Table Mountain is shown in background. View is to the southwest. Photo by Jon White.
Other examples of sinkholes and subsidence related to evaporite rock dissolution can also be seen in the report for the new CGS on-line publication, Colorado Map of Potential Evaporite Dissolution and Evaporite Karst Subsidence Hazards