Department of Transportation

Drilled Shafts Guidelines

Drilled shafts are used to resist relatively high lateral loads. They can be more economical than a spread footing/pile cap when a large cofferdam would be needed. Drilled shafts are sometimes selected to satisfy hydraulic or environmental footprint issues. Drilled shafts can provide good resistance to lateral soil movement and can be used to arrest landslides, when practical.

All drilled shafts are to be socketed into bedrock in West Virginia. The top of the rock socket should start near the rock-line or at the scour depth, if deeper. The rock socket is usually 6 inches smaller in diameter than the shaft portion to form a ledge on which to better seal the casing. For shafts at known dry-socket locations, the 6-inch step can be eliminated. It is preferred for the top of the shaft to be placed at least 1 foot above the normal pool elevation to facilitate latent concrete removal and to provide the joint into the cap above the normal water surface.

Initially, drilled shaft rock sockets should be sized to limit lateral displacement and geotechnical fixity in the service limit state. The longitudinal reinforcement should be selected in the strength limit state to provide a realistic flexural stiffness for the shaft for this lateral analysis. The socket depth and geotechnical fixity required by the WVDOH, is where the deflection curve crosses the zero line for the second time in the service limit state using soil-structure modeling software such as LPile. When using LPile, the internally generated P-Y curves for Strong Rock (Vuggy Limestone) and Weak Rock (Reese) should be used.

The internal P-Y curve for strong rock is when the unconfined compressive strength (UCS) is 1,000 psi or greater. To ensure deflection is limited to the initial part of the P-Y curve, the LPile Technical Manual limits deflection to 0.0004b at the top of rock (b = diameter). It is recommended that a lateral load test be performed in the field if the computed shaft deflection is more than the above deflection limit. For weak rock, we allow a UCS of up to 1,000 psi, even though a warning is issued over 500 psi. The input Krm for weak rock ranges from 0.0005 to 0.00005 for stiffer to softer rock, respectively.

Alternative computer methods for determining loading for shaft sizing and establishing socket depth may be used with the approval of the WVDOH. These methods may include the restraint from a group of shafts acting through a cap and/or the restraint from the bridge acting through bearing to other foundations.

The deflection needs to be checked in the service limit state. The amount of acceptable deflection is to be determined by the structural engineer. The deflection must be extended to the top of the substructure or bearing.

For large bridges, the vertical elastic response of the substructure and rock must be checked to determine the amount of settlement in the service limit state. The 2012 edition of LRFD is to be used to determine settlement of rock and the rock mass modulus Em using either the RMR method or the presented tables.

Cyclic loading from wind and water loading should be used when relying on the soil column to provide some of the lateral resistance, but no effect on rock is observed from cyclic loading in LPile. For seasonal cyclic loading of bridges due to thermal movement, we suggest 600 cycles for the 75-year service life. For high mast towers, 3,000 cycles are more appropriate from wind loading.

Normal factored loads are used to determine maximum moments and the average shear at the top of rock for structural design. The shear is averaged over one shaft diameter within the shear spike. Again, the geotechnical properties are not reduced in the strength limit state since this can lead to unrealistic response to loading. The axial geotechnical resistance is also checked in the strength limit state.

The axial geotechnical resistance of drilled shafts must be designed in accordance with LRFD, using the RMR system to select the bearing resistance equations as discussed for spread foundations on this webpage. A resistance factor of 0.45 must be used to determine the factored end bearing resistance, and a resistance factor of 0.55 must be used to determine the factored side resistance. When bedrock is dipping between 20° and 80°, the resistance factors for both end bearing and side resistance should be reduced by 0.50 to account for loss of resistance from the inclined bedding. The shaft side resistance must be calculated using Equation of the LRFD Bridge Design Specifications, 2012. Neither end bearing nor shaft resistance should be estimated per the latest edition of the LRFD Bridge Design Specifications for drilled shaft foundations. For non-redundant shafts, the factored resistances must be further reduced by 20%.

Claystone, mudstone, and Intermediate Geomaterials (IGM) may be prone to smearing and rapid deterioration upon exposure to air, water, or slurry, and therefore the estimated side resistance of the materials cannot normally be relied upon. However, if load testing is performed on a representative shaft(s), side resistance of claystone, mudstone, and/or IGM can be determined and the geotechnical engineer can judge whether the side resistance can be relied upon considering the results of the load test, the depth of the strata, and construction procedures of the test/production shafts. Alternatively, jar slake testing can be performed on core samples and if the material does not “flake” in 96 hours (per WVDOH Standard Specification 625.5.2.3 for maximum rock socket exposure time), it can be judged by the geotechnical engineer as reliable. Refer to the following link for IGM for a more detail definition of these materials. When a preliminary estimate of skin friction for these materials is needed, the formulas presented in FHWA-NHI-10-016, “Drilled Shaft: Construction Procedures and LRFD Bridge Design Specifications” should be used for IGM and not Equation referenced above.

Typically for rock sockets in West Virginia, both side and end bearing resistances can be added together directly or as discussed in FHWA-NHI-10-016. The exception is where the side-wall portion of the socket is medium hard or harder rock and the rock in the base is soft. In this case, the base resistance may not be mobilized due to the load-displacement behavior of the shaft. End bearing on rock usually provides more resistance than shaft resistance and consequently shallower/larger diameter sockets are considered more economical than deeper sockets once geotechnical fixity is satisfied.

When preinstallation borings are needed to verify the top of rock depth, Standard Penetration Testing (AASHTO T-203) is required at 5-foot intervals and core drilling should commence at either the earliest opportunity or at least at 50 blows per 3 inches. When augering in intermediate geomaterial (IGM), 2.5-foot sampling intervals are recommended. Do not power auger into rock. The top of rock socket elevation should be verified by either the geotechnical engineer of record or the WVDOH Geotechnical Group from the driller’s logs with the ground elevations shown on the logs. When end bearing is relied upon, the strata two shaft diameters below the tip elevation should be penetrated with the preinstallation core borings. Otherwise, one diameter is adequate. The rock within the two diameters should be free of voids and rock softer than anticipated. The base of the shaft should not rest on coal or gouge, but may bear on soft rock, if so designed. If the preinstallation borings encountered these bottom conditions, the core hole should be extended to obtain the full number of diameters beneath the tip.

A full-size nonproduction demonstration shaft should be performed for each project with shafts larger than 9 feet in diameter that are constructed in water. Cross-hole sonic logging (CSL) tubes are to be installed in all demonstration shafts and all wet shafts. Demonstration shafts are to be constructed using the same equipment, timing, and techniques as the production shafts. The reinforcement cage above the rock socket portion of the demonstration shaft can have a reduced number of bars and hoops provided the cage is strong enough to support the CSL tubes during installation. If the CSL results indicate a defect, another demonstration shaft may be required by the WVDOH. A written corrective action plan must be provided by the contractor discussing the problem(s) and its remedy. Normally, the demonstration shaft must be removed to 2 feet below the mudline.

For rating CSL results, both the increases in the First Arrival Time (FAT) and the energy reduction relative to the FAT or energy in a nearby zone of good concrete must be considered. The criteria for rating the concrete from the CSL test shall be:

Rating Criteria
Good (G) FAT increases 0-10% and Energy Reduction < 6 db
Questionable (Q) FAT increases 11-20% and Energy Reduction < 9 db
Poor / Flaw (P/F) FAT increases 21 to 30% or Energy Reduction of 9 to 12 db
Poor / Defect (P/D) FAT increases 31% or more or Energy Reduction >12 db

The flaw or defect zones and their horizontal and vertical extent must be reviewed by the geotechnical engineering of record or the Geotechnical Group. Flaws must be addressed if they affect more than 50% of the tested tube pairs at the same depth. Defects must be addressed if they affect two or more of the tested tube pairs at the same depth. At a minimum, addressing flaws and defects must include recording offset CSL measurements between all tube pair combinations where a flaw or defect was identified, as well as cross-hole tomography (CT) using all offset data. Core drilling and sampling may also be required for further evaluation of the flaw or defect, if identified based on the results of CT. The number, locations, diameters, and depths of the core holes and lengths of individual core runs will be determined by the geotechnical engineer or CSL/CT testing company and approved by the WVDOH Geotechnical Group prior to implementation. Coring procedures must minimize erosion of the core samples, avoid damage to the steel reinforcement, and be in accordance with Section 625.2.6.6 of the WVDOH Standard Specifications.

Bi-direction (O-cell) or lateral load tests are to be performed in the field when either the potential savings from reducing the size/length of the shaft offsets the cost of the test, or when required by the geotechnical engineer to verify the geotechnical resistances. The test shaft must be at least 80% of the production shaft diameter for dependable scalability. The bi-directional test shaft can be designed to test both end and side resistance or just the resistance in question. The test should fail the shaft or reach at least 150% of the nominal maximum load. Failure is normally considered as displacement of 4% of the diameter. Failure of either the socket rock, or concrete, or the equipment will be considered as the nominal resistance. The test shaft is normally removed to 2 feet below the mudline.

For shafts at a site that are represented by the test shaft, the compressive resistance factor can be increased to 0.7 (both side and end bearing) and for uplift (side only), the resistance factor can be increased to 0.6. Otherwise, use 0.45 end bearing and 0.55 side resistance.

The top of the shaft can have a maximum horizontal deviation of 3 inches from its plan location. The FHWA-NHI-10-016 manual recommends “including the axial loads applied at the bounds of the eccentricity permitted by the construction specification.” This 3-inch permitted eccentricity can result in a significant moment and must be considered in the strength limit state when sizing the shaft and reinforcement. For LPile, adding the moment from eccentricity to the pile head moment includes it into the P-delta effect. The shaft plumbness must be within 1.5% (0.15 feet per 10 feet of shaft), but normally is not included into the eccentricity. The reinforcement cage can deviate within the cap 3 inches higher or 1 inch lower than plan elevation when considering development length of the bars and top cover in the cap. If a shaft is constructed outside of the construction tolerances, it must be analyzed and remediated to meet the design requirements.

Since end bearing is typically used in West Virginia, rock socket bottom cleanliness is very important. The WVDOH requires mini-SID or SQUID inspection results to meet the following standard for cleanliness and that no more than 2 hours have elapsed between the inspection and startup of concrete placement. The base of the rock socket is to be considered clean if no more than ½ inch depth of material exists on 75% of the bottom and the remaining 25% has no particles or material more than 1-½ inches in thickness. Otherwise, the socket needs to be recleaned prior to placing concrete.