H2GEO: Geotechnical Engineering for Water Resources

The results of engineering research in Mesa Verde National Park tells a rich story of diligent and industrious early Americans in their planning, building, and maintaining community reservoirs over long periods of time. Field evidence shows that these Indians had knowledge of hydrological and hydraulic principles; they harvested water where modern engineers would say there is none and moved the water into reservoirs under difficult conditions. It was concluded that the early people of Mesa Verde had good organization, a social structure, and were adept at technology transfer from generation to generation and from valley bottom to mesa top. The descendants of the Mesa Verde people can be proud of their technological heritage.

Guanella Dam is a new 34-foot (10-meter) high, zoned earthfill embankment dam completed in the fall of 2003 and put into service in 2004. The dam was designed and constructed using a fast-track approach to meet water needs of the City of Golden. The dam is located near Empire, CO at the site of a previous aggregate quarry operation. The off-channel reservoir created by the dam will impound a water supply of about 2,140 acre-feet (2.64 million cubic meters). Several unique site conditions presented interesting geotechnical challenges that had to be overcome to develop a successful project. The site stratigraphy varied significantly over short distances, and the dam alignment was carefully selected to take advantage of a high ridge in a low permeability silt layer. On-site soils had to be carefully evaluated to allow for economic development of embankment materials. Because ground water was very shallow, a soil-bentonite seepage barrier wall through a foundation sand and gravel layer was selected to connect the embankment Zone 1 core to the foundation, silt layer. The barrier wall was constructed using slurry trench methods. Finally, the highly pervious sand and gravel deposits in the quarry area extended away from the reservoir, beneath the adjacent creek. The seepage barrier wall beneath the dam had to be extended across the adjacent creek valley through these deposits to limit reservoir leakage. However, to allow for expedited permits, the valley cutoff wall had to include two "windows" to avoid working in the creek or in the embankment of adjacent US Highway 40. Seepage analyses were completed to verify that the partial valley cutoff wall would produce acceptable leakage rates.

This paper is a case study in the difficulties of providing satisfactory foundation seepage control in glacial foundations. Washakie Dam is an embankment dam constructed on glacial moraine and glacial outwash foundations. Heavy seepage had been observed at the 62-foot (19-meter) high embankment dam ever since first-filling. Initial dam safety modifications were completed in 1999. These modifications included a new sloping sand chimney drain and a new deep sand and gravel toe drain intended to address embankment and foundation seepage. When the reservoir was initially refilled in 2000, the seepage modifications did not perform as expected. Piezometric pressures in the embankment and foundation built up to higher levels than expected, and a serious seepage incident occurred during investigations completed at the toe of the dam. Evaluation of the performance of the dam during refilling indicated that the pervious sand chimney was actually blocking flow from one or more areas of very concentrated seepage, resulting in very high piezometric levels upstream of the sand chimney. Washakie Dam was further modified in the fall of 2000 to include a gravel drain layer downstream of the sand chimney. The dam has performed well subsequent to the additional modifications, with measured seepage flows of almost 600 gpm (2,300 liters per minute) with a full reservoir.

A 60-foot (18-meter) downstream raise to the Loveland Water Storage Dam (aka Green Ridge Glade Dam) was constructed in 2003–2004. Due to increased head associated with an enlarged reservoir, the design included installation of a double-row foundation grout curtain. During construction, the field engineering staff responsible for construction oversight utilized an innovative approach to track and manage the foundation grouting operations. Information about each grout hole was entered into AutoCAD, with pertinent data separated by layers within the electronic file. Layers within AutoCAD were then be isolated or combined to generate various views of the overall grout curtain. For example, primary, secondary, or tertiary holes could be isolated to assess progressive closure. Alternatively, the entire curtain could be viewed and zones that accepted significant volumes of grout evaluated against the subsurface geology. Updated daily, this approach provided near real-time analysis of the grouting progress and allowed for rapid implementation of changes to the grout curtain design as necessary. Additional benefits over traditional methods of on-site grouting recordkeeping included easy transfer of data to the design team and others via electronic mail, and the effective completion of an as-built record in the field.

During the original construction of Pueblo Dam, a deep plunge pool was excavated at the toe of the massive head buttresses forming the spillway section of the dam. This section is founded on horizontally bedded sandstone and shale. While the original design engineers recognized the presence of a thick shale bed under the right-most spillway buttresses, and took measures to treat it, many thin shale beds and a continuous open sandstone parting went untreated. Dam safety evaluations indicated the potential for high risk due to sliding on the untreated planes daylighting in the plunge pool, if the reservoir exceeded the historical maximum level. To reduce this risk, the plunge pool was filled with roller-compacted concrete, and rock bolts were installed. Installation of a reinforced concrete slab and impact blocks on top of the RCC resulted in improved hydraulics for the spillway.

The Colorado Front Range of the Rocky Mountains was ravaged by wildfires during the early summer of 2002. Especially hard hit were areas in the watershed surrounding Cheesman Reservoir, which is located in the South Platter River basin in the foothills above Denver. The conditions left behind by a wildfire of this nature are susceptible to catastrophic mudflows, especially in mountainous regions, Denver Water became concerned that sediment from these mudflows could potentially impact Cheesman Reservoir. URS Corporation was employed by Denver Water to develop alternatives to minimize sediment deposition in Cheesman Reservoir.

Dry Creek is a north branch tributary to the Cache la Poudre River located near Fort Collins, Colorado. The Dry Creek upland basin is long and narrow with its source located in the foothills of the Rocky Mountains. Seven major irrigation canals and fourteen storage reservoirs form a complex hydrologic and hydraulic system in the upper basin. The lower basin has been developed in recent years and, as such, the capacities of the existing Dry Creek channel and floodplain have been greatly reduced, creating a significant flood hazard for the urbanized downstream area. For peak discharges in excess of the approximately 18-year frequency event, the Larimer and Weld Canal will overtop and cause significant amounts of flooding downstream. During the 100-year frequency flood event, over 1,600 residential and commercial structures would be impacted. The main goals and objectives of the project include: (1) protecting life and property within the developed area of the City of Fort Collins and (2) minimizing or eliminating the Federal Emergency Management Agency (FEMA) regulatory floodplain within the City of Fort Collins. Following an alternatives assessment that incorporated this unique hydrologic and hydraulic features of the basin, a combined alternative that includes detention, conveyance and diversion was selected for final design.

This article uses a recent study of riverbank erosion to showcase some of the conceptual features of evaluating an eroding soil slope. The focus of this article is on the assumptions and decisions made before and after the stability calculation. Although the larger project evaluated the overall impacts of the river on erosion, this article will only discuss the riverbank stability component. The choices made prior to the calculation, including identification of the input parameters and analysis methods, determine the calculation output and needs to be considered in the output interpretation. Evaluating natural, non-engineered slopes also requires careful evaluation of safety factor concepts. Failure types and the environments they result from are likely more significant for the understanding of erosion potential than the safety factor.

The City of Grand Forks, North Dakota is located along the Red River near its confluence with the Red Lake River. In April 1997, a 210-year flood event overtopped the levee system, resulting in the evacuation of the city and subsequent damages to residential and commercial structures and infrastructure exceeding $1 billion. Following this event, the U.S. Army Corps of Engineers (COE) began designing a higher flood control system, including levees and floodwalls, to contain future flood events. Because of the presence of deep, weak soil deposits along the riverbanks, analyses indicated that the system (primarily levees) would need to be setback from their existing location to achieve global slope stability, which would result in the taking of over 400 structures (mainly residential homes). As a result, the City of Grand Forks (City) retained Shannon & Wilson, Inc. (S&W) to explore alternatives to the setback levee system at four reaches along the alignment. This paper discusses the alternatives analyzed; the properties of the weak soil deposits; design methodology and criteria; the results of limit equilibrium and soil-structure interaction stability analyses; and the final design and construction of one alternative (a concrete cylinder pile wall) along a key river reach.

The St. John's Tunnel, constructed in 1903, is located on the south slope of Pikes Peak, Colorado at over 3,350 m (11,000 feet) elevation. The majority of the tunnel is unlined and, because of past instability, houses a 0.76-m (30-inch) -diameter steel pipe for conveyance. The tunnel is a critical element of the water supply system for Colorado Springs Utilities (CSU). Because of past rockfall and its future potential, CSU was concerned about the long-term reliability and integrity of water conveyance through the tunnel. In the summer of 2003, CSU let a design-build contract to evaluate and rehabilitate the tunnel support problems. Because of difficult underground access, operational constraints, and the necessity to complete the work prior to inclement weather, the successful completion of the project required the adoption of a number of innovative construction methods. This paper describes tunnel geology, ground support recommendations (short mechanical bolts and steel sets), prioritization of improvements, and the unique scaling and rock bolting stabilization program undertaken.

The remediation of the Burlington Mine Site, an inactive fluorspar mine located in the foothills of Boulder County, near the Town of Jamestown, Colorado, incorporated many innovative techniques and relied on geologic and geotechnical investigations, including a seismic refraction survey, drilling, and tomographic imaging. Innovative design treatments exceeded basic closure requirements, protected closures long-term, prevented future collapse within the mine workings, imitated natural form and function, and enhanced wildlife habitat. Treatments were required to meet multiple project goals and accommodate the abundant constraints and challenges at the mine site. Innovations included measures to address drainage issues at pit and adit closures, channel design to imitate natural form and function, waterfowl protection, and a specialized revegetation plan to restore riparian habitat and enhance habitat value on the site for future wildlife usage. These measures are important improvements over more traditional mine site remediation, which has always addressed safe conveyance of onsite water, but not necessarily how to restore natural form and function or how to use the water to maximize habitat value for wildlife. Exceeding minimum requirements resulted in a remediated mine site with more natural and ecologically functional systems that help soften the harsh aesthetics of the old mine site and improve wildlife habitat.

The state of the art in geotechnical modeling has reached the point where three-dimensional models of geologic and engineered structures can achieve accurate results for feature sizes ranging over several orders of magnitude. This paper presents the results of one such model. The Waste Shaft Station is one of the most critical openings at the Waste Isolation Pilot Plant (WIPP) underground facility. All transuranic waste that will be permanently disposed of at WIPP must pass through the station before disposal in the panels. It is one of the largest and oldest openings at WIPP and is expected to serve its current function for approximately 30 more years. To support efficient and safe operations in the Waste Shaft Station, a large FLAC3D model was developed and analyzed. The model encompasses the station itself and the drifts and pillars immediately surrounding it, an area of nearly two square kilometers. The model was also required to provide reliable results for excavation features on the scale of one meter. Convergence data from the station and from the adjacent drifts was compared to the model results.

The Pinery Dam (a.k.a. Bingham Lake Dam), located near Parker, Colorado, is a zoned embankment dam, over 50 feet (15.24 meters) high, constructed in the early 1970's. The Class 2 dam has a 350-foot (106.68-meter) long, 18-inch (45.7-cm) diameter, RCP (reinforced concrete pipe) outlet conduit with a reinforced concrete cradle. Leakage flows coming out of the existing outlet conduit were continuous and measured over 60 gpm (gallons per minute) (227 lpm (liters per minute)) with small amounts of embankment material. The source of the leakage problem became apparent when a remote-controlled video camera was able to view a 3-inch (7.6-cm) vertical offset at a pipe joint about 260 feet (79.25 meters) from the downstream end of the outlet conduit. Since draining the reservoir was to be avoided, any repair solution had to be performed either underwater on the upstream side or entirely from the downstream end of the outlet conduit. In 2002, a failed repair attempt was made using an inflated flexible-tube type lining set in place with epoxy resin from the downstream end. During this repair installation, the leakage flows and pressures caused the flexible-tube lining to be forced back out from the outlet conduit, ending in the repair failure. Several other repair alternatives were considered, but the final repair using a HDPE (high density polyethylene pipe) slip lining with grouting was selected. The HDPE slip lining solution was the least expensive alternative, but was risky, because of the leakage flows and the 260-foot (79.25-meter) distance to the 3-inch (7.6-cm) vertical offset in the 18-inch (45.7-cm) RCP. This selected repair solution required the use of an inflatable packer, remote-controlled video camera, and 12, 1-inch (2.54-cm) diameter grout lines at up to 300 feet (91.4 meters) long. In the annular space between the 14-inch (35.56-cm) HDPE and 18-inch (45.7cm) RCP, 12 individual grout cells were filled with non-shrink cementitious grout along the entire slip lining length. After the successful HDPE slip lining installation in 2003, the measured leakage flow was reduced to about 4.8 gpm (18.17 lpm) from the repaired outlet conduit.

In 2002, the City of Brighton needed to expand the capacity of its South Platte Alluvial Aquifer wellfield. Simultaneously, a drought was occurring that was causing the water table in the wellfield to decline. We were asked to find a way to quickly add a well to the existing wellfield that could meet the increase in demand (900 gpm) in the face of declining water levels. Based upon our analysis and the results of a test drilling program, we implemented four necessary strategies. To minimize permitting and construction time, we located the well on City property. To minimize interference caused by other wells, we found a well location that maximized the distance from existing wells and also met our minimum saturated thickness criterion. To meet the demand, we utilized all of the available drawdown by setting the pump inlet below the bottom of the well screen in a sump. To avoid air entrainment in the water when the well was operated at maximum drawdown, we increased the well diameter and dramatically increased the well screen slot size and gravel pack size to three times that traditionally used. This prevented air entrainment by preventing turbulent flow thru the well screen. The resulting well can produce 1200 gpm with 11 feet of drawdown and has a specific capacity 1.2 times other wells in the field.

The design and construction of the soil-bentonite liner for the Hazeltine, Road Runners Rest II and Brinkmann-Woodward Gravel pits (Hazeltine) involved a 15- to 63-foot-deep by 3-foot-wide by about 3-mile-long cutoff trench filled with a soil-bentonite backfill that would combine three existing sand and gravel mines into one water storage facility that will ultimately provide Denver Water with 5,100 acre feet of storage along the South Platte River. The design and construction of the reservoir was accelerated to accommodate construction of 120^th Avenue. The new roadway crosses through the Brinkmann-Woodward and Road Runners Rest II pits, and requires up to 15 feet of fill to be placed above the soil-bentonite cutoff trench. The factors affecting liner construction included a tight schedule and ongoing mining operations that constantly changed site conditions throughout design and construction. The team of GEI Consultants, Inc. (GEI), Denver Water, and Envirocon, the contractor, quickly adapted an interactive approach to meet the challenges of construction at a site that changed daily, including modifying the alignment and placing over 200,000 CY of fill to create a working platform for construction of the trench. The team also had to respond to unforeseeable site conditions, like a 450-foot-long, 12-foot-deep section of concrete that had to be removed for liner construction. This paper discusses the construction challenges of the Hazeltine, Road Runners Rest II, and Brinkmann-Woodward gravel pit liner.

The Denver Basin aquifers are a widely developed ground water supply for many uses along the Front Range of Colorado, Wells are commonly constructed using reverse circulation mud rotary drilling methods. In the Denver Basin, prior to 2003, there were no incidents of major drilling fluid loss or lost circulation of properly conditioned drilling fluids resulting in the loss of the borehole. In 2003, three separate incidents of significant drilling fluid loss occurred within a sub-linear geographical area of approximately five miles in length in and near the City of Aurora. One of these incidents resulted in the loss of the borehole and the loss of the contractor's down-hole drilling equipment. In all four wells, the fluid loss occurred near the contact of the Cretaceous Fox Hills Sandstone and the overlying Cretaceous Laramie Formation. Fluid loss occurred in a highly permeable or fractured sandstone at depths of approximately 2,000 feet (610 meters). Despite the fluid loss zone encountered in these wells, many other wells have been completed without significant fluid loss in the same geographical area. Important lessons were learned as a result of these fluid loss incidents regarding the importance of drilling fluid volume, mud program control, reaction time in the case of a fluid loss incident, and awareness of the potential for fluid loss incidents to occur in the Denver Basin.

The Morrison Quarry Reservoir No. 1 is a unique water storage reservoir located west of Denver in the town of Morrison, Colorado, The reservoir was created as a result of open-pit mining of crushed stone aggregates at Aggregate Industries' Morrison Quarry and provides 620,000 m³ (500 acre-feet) of raw water storage for the Town of Morrison. The quarry provides a good example of public/private partnering between Aggregate Industries and the Town of Morrison, where a private aggregate producer partnered with a public entity to provide a needed water storage reservoir and reduce long term closure costs for an existing aggregate mine facility. Additional benefits of the project include the fact that the quarry and reservoir project provides needed aggregates for Denver area construction, and benefits the environment by utilizing existing mined lands and therefore avoids the need for other reservoirs in more environmentally sensitive areas. Morrison Quarry Reservoir No. 1 is the first mined rock quarry to be used for water storage in the Denver-Metro area. This paper presents the project through the perspective of the geotechnical engineers, focusing on geologic conditions and technical issues important for development and operation of the reservoir. The key technical issues for the water storage reservoir development included the potential for water seepage from the reservoir through the jointed and fractured rock mass and the stability of the rock quarry slopes.