Laserfiche WebLink
Western Dam Engineering <br /> Technical Note <br /> <br /> May 2016 <br /> <br />26 <br />no piezometers exist to provide an accurate <br />understanding of the foundation pressures, field <br />investigations to install piezometers may be <br />warranted. A series of piezometers screened to <br />measure pore pressures at different elevations and <br />locations is required to estimate seepage gradients. <br />Seepage gradients, along with the foundation material <br />type, will control the potential to initiate internal <br />erosion. As previously discussed, several methods are <br />available to estimate critical seepage gradients and all <br />cautions previously stated for these methods apply [5], <br />[6], [7], [8]. <br />If piezometers are to be installed, care and proper <br />design are required to strategically select their <br />locations and installation methods. This will help to <br />avoid initiating a dam safety incident due to drilling, <br />yet provide instruments that can capture the desired <br />information. See the section Don’t Rush to Dig In for <br />more information. An example of piezometers located <br />to measure seepage pressures within the pervious <br />aquifer and overlying confining layer is shown on <br />Figure 15. Piezometers can be installed relatively <br />inexpensively within small dams using truck or trailer <br />mounted drill rigs. Cone Penetration Test (CPT) rigs <br />can also be a cost-effective option for installing fully <br />grouted piezometers while limiting disturbance. <br /> <br />Figure 15. Piezometers Located to Measure Potential <br />Uplift <br />A rough rule of thumb is to assume a safety factor of 2 <br />when the piezometric surface in the pervious water- <br />bearing strata is at ground surface, based on the unit <br />weight of soil typically being nearly twice the unit <br />weight of water. Higher piezometric levels in the <br />water-bearing strata would result in a reduced factor <br />of safety. <br />Seepage weirs can also provide a means of monitoring <br />for increase seepage flows or sediment accumulation. <br />See the previous Western Dam Engineering Technical <br />Note article for more information on dam safety <br />instrumentation: Does Your Dam Measure Up? <br />Numeric Seepage Analysis <br />A qualitative analysis using field observations and <br />instrumentation data can sometimes provide enough <br />information to understand the problem and guide the <br />engineer on what to do, and just as importantly, what <br />not to do. However, if the existing knowledge and <br />instrumentation data are not conclusive, then a two- <br />dimensional model of the dam and foundation can be <br />analyzed using appropriate stratigraphy and material <br />properties. <br />Seepage analyses can be performed either by <br />computer software (which is typical today) or by <br />manual flow net generation. The results of the seepage <br />model can be used to develop a better understanding <br />of the seepage regime beneath and through the dam <br />and estimate foundation pressures and hydraulic <br />gradients. Calculated gradients can then be evaluated <br />for the potential to initiate erosion. Typical software <br />codes, like Seep/W by Geostudio [3], are relatively <br />easy to learn. However, the model results are only as <br />reliable as the model inputs. If the variation in <br />stratigraphy and material properties is not well <br />defined, then spending time and money to develop a <br />model is a pointless exercise. Before investing in <br />developing a seepage model, first do a reality check by <br />looking at the site configuration and piezometric data <br />(if available) and deciding whether a two-dimensional <br />model would adequately represent the seepage <br />regime. It is common for dams and foundations to <br />have seepage flow (gradient) that is not perpendicular <br />to the dam axis. If this is the case, it can be difficult to <br />develop and calibrate a meaningful two-dimensional <br />model. Three-dimensional models have a higher cost, <br />are more susceptible to inaccuracies, especially with <br />limited data, and require special expertise. All of this <br />needs to be considered when determining the most <br />appropriate, and justified, means of evaluation. <br />