U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Bridges & Structures


Interstate Technical Group on Abandoned Underground Mines
Fourth Biennial Abandoned Underground Mine Workshop

Real Time Monitoring of Subsidence Along Interstate I-77, Summit County, Ohio


Kevin O'Connor GeoTDR, Westerville, Ohio
Rick Ruegsegger Ohio DOT, Office of Geotechnical Engineering, Columbus, Ohio
Kirk Beach Ohio DOT, Office of Geotechnical Engineering, Columbus, Ohio


The Ohio Department of Transportation (Ohio DOT) encountered abandoned underground coal mine subsidence during an investigation conducted as part of a project to widen I-77 in Summit County, Ohio. A decision was made to stabilize 2000 feet of four-lane highway by backfilling the mine voids with cement grout while traffic remained on the interstate. Due to the potential for additional mine subsidence during remediation work, Ohio DOT required the installation of areal-time monitoring system to activate an alarm when movement or settlement of the road base exceeded threshold values. Innovative monitoring of ground deformation was accomplished using time domain reflectometry (TDR). This technology is a form of RADAR in which voltage pulses are transmitted along coaxial cable and reflections are created at every location where the cable is being deformed. The distance to each location is determined by the pulse travel time, and the magnitude of deformation is determined by the magnitude of the TDR reflection. This allows for real time monitoring of all locations where ground movement is causing deformation of the cable, and the rate of deformation at each location. The cables were installed by a directional drilling contractor in horizontal holes drilled 5 feet or more beneath each lane of the entire section of highway. The cables were connected to a central remote data acquisition system that automatically recorded and stored measurements. The initial record for each cable was stored as the baseline measurement. When the magnitude of any TDR reflection differed from the baseline value by a specified amount, the system initiated a phone call to GeoTDR personnel who could then analyze TDR measurements in real time via a phone line connection to verify that the alarm was associated with cable deformation. If it was determined that actual movement of soil had occurred, Ohio DOT personnel were alerted to intensify visual reconnaissance and determine if lane closures were necessary. This project demonstrated the capability of real time monitoring of ground movement over a wide area utilizing a horizontal application of TDR technology.


The pits, sags, and troughs that develop on the road surface over abandoned underground mines are the ultimate result of settlement in subsurface strata. Subsidence characterization of the ground surface over abandoned room-and-pillar mines remains relatively undeveloped due to the complexity of geologic conditions and the variability of time-dependent rock mass behavior at the mine level and within its overburden. In the case of highways located over abandoned mines, it is not enough to say that subsidence is imminent. The progression of a subsidence event from the mine level to the ground surface is guaranteed. When will it happen? Will the event be sudden or gradual? Will the event result in a sag in the road or a sudden shear. A means must be provided to detect strata movements and quantify the rate at which they are occurring so that appropriate measures can be taken to protect the traveling public and mitigate damage. One component of the array of techniques available has involved monitoring of subsurface movements at depth which are precursors of surface subsidence.

Underground coal mining was active in Summit County, Ohio from 1810 to 1970, and the area is underlain by a network of mine voids. Support for the overlying rock is provided by remnant pillars and blocks of coal as well as water pressure (in the case of flooded mines). As a result of pillar failures, roof rock failure, and changes in water pressure within the abandoned underground mines, there have been localized mine subsidence occurrences. Sinkholes such as the one shown in Figure 4 were detected during the site investigation as part of a project to widen I-77.

The project site is located in Summit County, Ohio (Figures 1 and 2) and has been extensively undermined for the Brookville #4 coal seam (5 to 6 feet in thickness). The underground mines are covered by a thin layer of bedrock that is overlain by a mantle of glacial deposits. Maps (Figure 3) were assembled and digitized for the Overholt Mine operated by Overholt Coal Co. until 1936, and the Haurer Mine operated by R&T Coal Co. until abandonment in 1937. These mines operated with an extraction ratio of over 50 percent.

Motivation Based on Past Experience

The most common method for stabilizing abandoned mines is to backfill with fly ash or cement grout (Figure 8). In the case of flooded mines, cement grout is used to prevent contamination of ground water. Ohio DOT gained considerable experience in 1995 when backfilling mines beneath Interstate Highway I-70 east of Cambridge in Guernsey County, Ohio (Willard, 1995).

The injection of grout into the flooded abandoned mines at a depth of 40 feet beneath I-77 had displaced of water impounded in the mine works (Figure 5) and had induced subsidence events in the project vicinity. Since a condition for the Summit County I-77 project construction was maintaining all lanes available for traffic through the project area, the potential for inducing subsidence had to be addressed.

Plan of Action

The plan of action developed by Ohio DOT was multifaceted with the primary objective being protection of the driving public. Proactive components of the plan included installation of an alarm system, visual monitoring, and backfilling the mine. Reactive components of the plan included reducing the speed limit to 50 mph, lane closures, and provisions to detour traffic in case closure of all lanes was deemed necessary. The alarm system consisted of TDR monitoring cables installed beneath the highway and connected to a central data acquisition system for automated monitoring. Complementing the TDR monitoring cables were geophysical surveys in areas of suspected mine shafts.


The instrumentation was installed to provide a real time monitoring and alarm system. Precursor subsurface deformation was monitored by installing coaxial cables, as shown in Figure 6, into holes drilled horizontally beneath the highway as shown in Figure 7. Ohio DOT required a system that would monitor settlement beneath every lane open to traffic. The original system plan called for installation of cables in two trenches along the highway. Problems with installation and repair of the road surface deemed it necessary to consider other alternatives. Subsequently, the final system configuration included three cables (TDR1, TDR2, and TDR4) installed beneath the existing pavement (Figure 9) and one cable (TDR3) installed in a trench that was ultimately covered with asphalt pavement for a temporary lane (Figure 12).

Alarm System and Visual Monitoring

A critical requirement for the monitoring system was an automatic, datalogger-initiated capability to alert Ohio DOT personnel in the event that precursor subsurface movement was occurring as mine grouting was conducted. Automation was accomplished by connecting the cables to a central data acquisition system controlled by a Campbell Scientific CR10X datalogger (Figures 13 and 15).

Once the unit was installed and operational, the sequencing of the alarm system required an initial base-line reading. Alarm limits were then established to regulate the degree of acceptable variation for each TDR cable. If at any time, these limits would be exceeded, the alarm would trigger an automatic, datalogger-initiated phone call to key personnel informing them where movement was detected while enabling real time monitoring via phone line. Based on this information, they could make a decision about alerting Ohio DOT personnel at the site who would make a visual inspection to determine if lane closures were warranted.

Principle of TDR

Innovative monitoring of ground deformation was accomplished using time domain reflectometry (TDR). This technology is a form of RADAR in which voltage pulses are transmitted along coaxial cable, and reflections are created at every location where the cable is being deformed (Figure 6). The distance to each location is determined by the pulse travel time, and the magnitude of deformation at each location is determined by the magnitude of its TDR reflection.

Prior to installation, the cable is crimped to provide reference reflections at known physical locations along the cable (Figure 9D). In vertical installations, the crimped cable is lowered down a borehole and typically bonded to the surrounding rock with an expansive cement grout that is tremied into the hole. At locations where progressive ground movement is sufficient to fracture the grout, cable deformation occurs that can be monitored with a TDR cable tester(O'Connor and Dowding, 1999). This technology has been effective for monitoring subsurface movement over abandoned mines (O'Connor and Murphy, 1995; Charette, 1993; Bauer et al,1991; Aston and Charette, 1993). The installation in Summit County was unique since it was the first project in which TDR monitoring cables were installed in directionally drilled horizontal holes. Furthermore, this project was the first application in which TDR was used to monitor movement over the lateral extent of an entire project area.

Installation Design

Initial Concept

The original monitoring plan entailed the installation of one cable in a trench along the shoulder, and a second coaxial cable within the pavement. The initial plan proposed a 2-inch wide groove to be cut down through the asphalt overlay into the concrete pavement for a total depth of 6 inches. The 7/8-inch diameter solid aluminum coaxial cable would be placed in the groove and backfilled. A non-shrinking backfill which would bond the cable into the pavement was required. Use of an epoxy backfill was not feasible since it required a 24 hour curing period and the lanes could only be closed from 7pm to 5am. Use of a hot mix asphalt was not feasible since its 200 deg-F temperature would melt the foam dielectric. A cold mix asphalt backfill was not considered feasible because compaction settlement of the asphalt backfill could create a residual groove in the driving surface that would be a problem for motorcycle traffic.

A test program would have been required to establish a viable installation procedure. This requirement was a problem due to the limited time available for cable installation before mine backfilling was initiated. Due to the difficulties with epoxy, grout and asphalt backfill, an alternative plan for cable installation was developed.

Horizontal Directional Drilling

The alternative cable installation involved horizontal directional drilling (HDD) beneath the pavement (Figures 7 and 9). The horizontal holes would be drilled from the midpoint of the 2000 ft long section of highway. Location and depth would be carefully controlled so that each hole would be beneath the centerline of a highway traffic lane at a depth of at least 5 feet, so as to avoid any subsurface drains. Holes TDR2 and TDR4 were added in conformance to the safety consideration that traffic should not travel in lanes which were not monitored or already stabilized.

This plan would have required a large directional drill rig, drilling an 8-inch diameter hole in order to have a continuous hole over a 1,000 feet length. Fortuitously, this large rig was not available and instead a smaller rig drilling a 4-inch diameter hole with a maximum hole length of 600 ft was utilized. This plan modification required that each 1000-ft-long section be divided into two 500-ft-long sections which were then spliced together (Figure 7).

Drilling was conducted at night when single traffic lane closures were permitted by Ohio DOT. The resulting access to the pavement surface above the drill bit would make it possible to provide the driller with information on the direction and depth of the drilling (Figure 9B). Consequently, horizontal and vertical control was not expected to be a concern even in the presence of reinforcement in the pavement. However, as discussed later, variation in the subsurface conditions created a challenge to the horizontal drilling operations.

Discussions with contractors revealed that the critical factors for HDD work were the length of pull, the tensile strength of the cable, the requirement of having good cable to soil contact, and the ability to accurately drill through sand and/or clay. The contractors recommended a pull length of less than 800 feet. The depth of the installation was intended to be 5 to 10 feet below the top of pavement to avoid disturbance to shoulder drains and storm water drains. With a shorter pull length, a smaller drilling rig could be used. To provide adequate clearance for the 7/8-inch diameter cable, a 3 to 4-1/2 inch diameter hole would be drilled with mud (bentonite slurry). The bentonite would help stabilize the sand layers while providing a lubricant to back-pull the cable which had a tensile strength of only 800 lbs.

The contractor proposed that grouting in the annulus between the cable and hole wall could be performed in several ways. One alternative involved grouting the annulus with a bentonite-neat cement mix during the back-pull. A second alternative involved placement of the cable inside a tremie line. Both the cable and the tremie line would be pulled back together. Once the pull was completed, the cable would be anchored from one end, and from the opposite end the tremie line would be connected to the grout pump for retreat grouting. All parties agreed that the best grouting method should be determined by field testing.


Subsurface Conditions

The subsurface conditions consisted of 15-30 ft of clayey to sandy silt containing layers of clean sand to gravelly sandy. These unconsolidated deposits were overlying 6 ft or less of broken and jointed limestone and the 5-6 ft thick Brookville #4 coal seam. The difficulty of drilling through these soils was demonstrated by vertical test drilling performed to establish appropriate procedures for production grouting operations in preparation for mine backfilling. It was found necessary to case all vertical holes through the cohesionless soil using a vibratory drill rig in order to minimize washing out of the soil and the risk of forming sinkholes.

The water table was encountered at a depth of 6-10 ft below the ground surface; therefore, the mine voids, located at a depth of 25 to 35 feet, were flooded. Communication between the vertically drilled mine backfilling holes (Figure 5) occurred within the gravelly sand and broken limestone as well as at mine level. These "soft zones" and the variation in the depth to top-of-rock posed a challenge for the horizontal directional drilling.


The original installation plan anticipated the horizontal drill being able to negotiate fairly tight horizontal curves (radius of 15 ft). However, the large horizontal drills actually required a larger radius curve (radius of 200 ft as shown in Figure 7) for negotiating through "soft zones." This large rig limitation would have created situations where 125 foot lane sections were without a monitoring cable. Fortunately, the smaller horizontal drill rig, shown in Figure 9, was used since it could work with a radius as small as 100 ft.

HDD requires a reaction force (resistance) at the drill head in order to control the horizontal and vertical position. Whenever it encounters weak soils or "soft zones," the bit will progress with increasing depth until it encounters sufficient resistance for the operator to turn the drill head and direct the bit upward. Conversely, when the bit encounters a hard object such as a boulder or a buried pipe, the bit will move laterally until it gets around the obstruction and will continue in this direction until it encounters a zone of sufficient resistance so that the operator can rotate the head and redirect the bit. Typically, the horizontal location of the bit could be controlled to within +/- 3 feet. The influence of subsurface conditions on horizontal control is illustrated in Figure 11 and Table 1. When drilling hole TDR1N-2, little or no resistance was encountered at Station 151+00 and this continued as the bit was advanced to Station 152+60. This low drill head resistance did not allow the operator to change either the horizontal or vertical direction of the bit. However, the operator was able to pull the bit back to Station 150+00 and redirect it.

Due to a problem with cable installation in TDR1N-2 (as discussed below), it was necessary to redrill this hole. As shown in Figure 11, the original hole was at an offset of 71.5 feet right of the highway centerline; and, it was redrilled with an offset of 74 feet right. Although this second hole was only 3.5 feet away from the original hole, the soft soils it encountered had a more dramatic impact on drilling. The bit encountered soft soil at a depth of 9 ft (at Station 147+50),and it dropped as it was advanced. It continued dropping to a depth of 16 ft (at Station 149+00) and could not be directed to a shallower depth until stiffer soil was encountered at Station 149+75.

By comparison, while drilling hole TDR2N-2 (offset 59 Right Figure 11) the soil was stiffer. This resistance made it easier to control the bit depth.

An unexpected benefit of HDD was the wealth of information it provided relative to soil conditions beneath the highway pavement. As summarized in Table 1 and Figure 10, it was possible to identify the location and lateral extent of soft soils. By comparison, it would require an extraordinary number of vertical exploration holes to provided a similar amount of soil information for the 2000 foot long section of roadway.

Cable Installation

At the end of a 500 ft run, the bit was directed to emerge along the pavement shoulder. The coaxial cable was attached to the drill string and pulled into the hole as shown in Figure 9D. The cable was then crimped every 45 ft when there was a pause to detach a section of drill rod and load it onto the rack. These crimps produced reflections at know distances in the TDR waveforms. An example of the reference reflections is shown in the waveform in Figure 16A.

Originally it was planned to anchor the cable to soil by pumping a lean cement grout through the drill stem as it was extracted. The grout mix was 600 gal water, 3 bags expansive additive, 15-17 bags of cement. This method was used successfully during the first hole (TDR1-N1). However, the mud pumps and drill-string back-reamer were not designed to be used with cement and plugging was problem. While installing the second cable (TDR2-N1) the back reamer became plugged and the mud pump broke down. Grout pumping was discontinued and the cable was simply pulled through the hole. At this point, the decision was made to discontinue attempts to grout the annulus to minimize delays that would be caused by changing equipment.

The lack of annulus grouting raised concerns that cable deformation would not occur. The 7/8-inch diameter cable was pulled into a 4-inch diameter hole filled with very soft drilling mud. However, a field observation suggested that the bore hole was collapsing on the cable. While back-pulling a cable through hole TDR1-N2, it was necessary to make an adjustment at the drill rig with 90 ft of rod still in the hole. The cable got stuck in hole and only 4 ft was pulled out at the drill rig. When an attempt was made to pull the cable out at the other end, it immediately snapped off. It was concluded that the soil had collapsed on the cable. This condition should occur for all cables installed in similar soil conditions, and any ground movement would cause cable deformation in these settings.

Installation in Trench

One cable (TDR3-N) was installed in a trench 1 ft wide by 3 ft deep as shown in Figure 12. This installation was very simple and controlled. The cable was laid out next to the trench, crimped, placed in the trench, then covered with 3 ft of sand-cement backfill (compressive strength 200 psi). Since the plan was to place a temporary lane over this trench, it was necessary to place controlled density cementitious backfill. This ensured that the backfill would not consolidate and cause settlement of the pavement. However, the usual procedure would be to only use 6 inches of cementitious backfill then complete backfilling with soil.

Automated Monitoring and Alarm Logic

The eight spliced coaxial cables extended to a central location (Figure 13, 14). Each cable was connected to the coaxial multiplexer installed within the enclosure shown in Figures 15. The multiplexer and Campbell Scientific TDR100 cable tester were controlled by a CR10X datalogger. The datalogger was also attached to a storage module and modem. Initially, there were difficulties in getting the automated system operational. During this period, all cables were interrogated manually using a TDR unit and a laptop computer. In this mode, the analysis was essentially qualitative with only a gross visual analysis to check if any cable deformation was occurring.

Once the system was operational, the datalogger would cycle over the eight cables once every three hours and store the measurements. The datalogger was programmed to compare waveforms point-by-point against stored baselines. The datalogger would activate power to the cable tester, interrogate each cable, compare TDR reflection magnitudes with the baseline waveform, store data in the storage module, and then turn off the cable tester. Whenever the system detected a change from the baseline value that exceeded the alarm threshold, the datalogger initiated a phone call to GeoTDR personnel. The alarm value was incrementally increased as personnel gained experience with the system. Data could be downloaded from the storage module via a phone line to evaluate the cause of the alarm condition.

The alarm logic is summarized by the flow chart shown in Figure 17. The memory locations available in the datalogger to store baseline waveforms placed a limitation on the number of data points that could be stored. For each cable, it was possible to store a maximum of 400 data points and the integrated length was standardized as 440 m (1443 ft) to ensure that the entire length of cable would be interrogated. This translates into a density of 400 points / 440 m (i.e., 1 point / 1.1 m = 1 point / 3.28 ft). The datalogger was programmed to compute the point-by-point difference between the current waveform and the baseline. That is, for data point "I", CurrentValue(I) - BaselineValue(I) = Change(I) (with I = 1 to 400). If the absolute value of"Change(I)" is equal to or greater than the threshold value of 0.1 rho (100 mrho), this would be an alarm condition, and the datalogger would initiate a call to GeoTDR.

Typical Data and Overburden Response

The monitoring system philosophy was to have an alarm condition if the cable was severed or if movement exceeded the established threshold values. However, it was more prudent to initially set the alarm threshold at a much lower level and then increase the threshold as experienced was gained. Examples of the types of cable faults that exceeded the lower alarm threshold, and associated waveforms, are as follows:

  1. Poor Connector (Figures 18A, C, E, G). Problems developed with connectors between lead cables and transducer cables, and with splices along the transducer cables.

  2. Cable damaged by Backhoe (Figure 18B). The end of cable TDR1-N was damaged by a backhoe digging in the area of Station 151+00. From that point in time, this end of cable was exposed and the outer conductor was damaged.

  3. Cable Severed by Drilling (Figures 18D, F, H). Since the system was intended to detect subsurface movement induced by the mine backfilling, it was necessary to have the cables installed prior to these activities. Cable locations were not well marked on the highway pavement and several drill holes intercepted the cables.

  4. Electrical Noise (Figures 18F, H). Sometimes electrical noise associated with poor connections was sufficient to exceed the alarm threshold.

Examples of ground movement that was not detected are as follows:

  1. Surface Movement not Detected in Soft Soils. The cable were installed at depths of 3 to 17 ft beneath the pavement. It is inherent that soil/rock movement would have to occur below this depth to cause deformation of the cable. At Station 149+50 (Figure 10 and 11), a small slump developed in the roadway early in the morning of August 3, 2001. An investigation of this slump by drilling and a geophysical survey showed that it was associated with "shallow" movement in soft soil and the grout take was very low. "Shallow" is defined as occurring above the cable that was at a depth of 9 feet at this location. Consequently, this movement within the soil was not detected by the cables, and it was not related to movement of the underlying rock.

    The presence of soft soils was also evident during horizontal drilling to install the cables. At Station 136+00, drilling fluid came out through cracks in the pavement (Figure 10).

  2. Cable Installed in Trench with Stiff Backfill (Figure 12). As mentioned, one cable was installed in a trench 3 ft deep with sand-cement backfill. On August 13, this "beam" was exposed along a 5 ft long section while repairs were made to a drain. The bending stiffness and shear strength of this "beam" was great enough that only elastic deformation occurred over the exposed span, cracking did not develop, and the cable was not deformed.

  3. Splice Hit by Automobile. On August 10, a northbound automobile went off the highway and hit the splices at Station 136+00 (Figure 7). The cables were bent gently and not locally deformed so this was not detected when interrogating the cable with TDR.

Limitations of Alarm Logic

The following information is a summary of calls that were logged and the analysis of the cause:

Date Location Comments
8/18/01      TDR4-N event at 9.9 m connector
8/22/01 TDR4-N event at 26 ft connector (Fig 18G)
8/29/01 TDR1-N event at 1158 ft      backhoe hit end of cable (Fig 18B)     
8/30/01 TDR4-N event at 195 ft cable severed by drilling (Fig 18H)
9/01/01 TDR1-N event at 1158 ft 3 received calls......
9/04/01 TDR1-N event at 1158 ft 2 received calls......
9/05/01 TDR1-N event at 1158 ft 4 received calls....
9/08/01 TDR2-S event at 29 ft splice (Fig 18E)
9/09/01 TDR2-S event at 29 ft continuation of alarm condition
9/20/01 TDR1-N and TDR2-S 8 received calls...
9/20/01 TDR1-S event at 14 ft connector (Fig 18C)
9/21/01 TDR1-N and TDR1-S 8 received calls...
9/23/01 TDR1-N and TDR1-S 20 received calls....
9/25/01 TDR2-N event at 14 ft connector
9/26/01 TDR2-N event at 18 ft connector
9/27/01 TDR1-S event at 155 ft cable severed by drilling (Fig 18D)
9/27/01 TDR2-S event at 837 ft cable severed by drilling (Fig 18F)
9/27/01 TDR1-S event at 155 ft continuation of alarm condition
9/27/01 TDR2-S event at 758 ft related to drilling
9/27/01 TDR2-S event at 404 ft noise level 100 mrho
9/28/01 TDR2-N event at 18 ft continuation of alarm condition
9/28/01 TDR1-S event at 155 ft continuation of alarm condition
9/29/01 TDR1-N and TDR2-S 12 received calls.....
9/30/01 TDR1-N and TDR2-S 12 received calls.....

The above summary documents the fact that once a cable is damaged, moisture can penetrate, and small fluctuations in the TDR waveform at that location will exceed the alarm threshold. This situation can also occur simply due to intermittent noise created at connectors.

Solutions used to Overcome Limitations

One solution to overcome these limitations was to simply have the system record the time that a call was received. This solution made it possible to identify a cable by the logged call time since cables are interrogated at specific times

Cable Time
TDR1-N    0005, 0305, 0605, 0905, 1205, 1505, 1805, 2105
TDR1-S 0015, 0315, 0615, 0915, 1215, 1515, 1815, 2115

A second solution was to reset the baseline. However, this solution was only a temporary fix since intermittent cable noise would recur and exceed the alarm threshold.

A third solution was to increase the "alarm threshold" from a "Change" of 0.1 rho (100 mrho) to 0.9 rho (900 mrho). This solution was effective in dealing with the noise created by the connector for cable TDR4N (Figure 18G).

Other Possible Solutions to Overcome Limitations

Several approaches could be used to address the current limitations. One approach would be to modify the basis for an alarm condition, such as : A) compare slopes for a series of points, B) compare the difference from the baseline for a series of points, and/or C) change the number of points / unit length. These modification should be made in concert with an evaluation of the mine geometry and the specific geologic conditions impacting the area.

An even more effective approach would be to dedicate a base station to download data automatically. This approach would be similar to systems used for web-based data displays since archived data would be more readily available for comparison. This would produce a more continuous time history of changes. Software could be utilized to detect and flag changes before an alarm threshold was exceeded. Most importantly, this approach would significantly decrease the likelihood of responsible personnel developing "burn-out." Several individuals could be involved in the process and a report of calls could be viewed daily rather than requiring individuals to be on call 24 hrs/day 7 days/week.

Summary and Conclusions

This project demonstrated the capability of real time monitoring of ground movement over a wide area utilizing a horizontal application of TDR technology and the viability of real time monitoring over abandoned underground mines. The following conclusions have been drawn based on the experience gained during this project:

  1. It is viable to install coaxial cable in horizontal holes for purposes of monitoring subsurface deformation using TDR technology.

  2. Horizontal drilling cannot be easily controlled when soft soil encountered, the bit enters a pipe,or the bit encounters hard obstructions. However, horizontal drilling does provide a good assessment of geologic changes over lateral extent of the project area. Perhaps a resolution to the problem with control in soft conditions could be addressed with a pilot bit designed specifically for these conditions although this would require delays to retract the drill string and change drill heads.

  3. Originally, two grouting methods were identified and it was planned that a field test would be used to determine which method was feasible. However, only one method was attempted, and the problems with grouting were created by the type of pump and drilling equipment used. Discussions with HDD equipment manufacturers and contractors indicate that this equipment is typically only used with bentonite slurries. It will be necessary to implement an alternative means for placing cement grout in horizontal drill holes. For example, a separate tremie tube can be placed with the cable, or a conduit can be placed as typically done when installing optical fiber. Grout would be pumped through the tremie tube or conduit and it would remain in the hole.

  4. Experience with the alarm was gained and as confidence increased it was possible to adjust the threshold values for individual cables. In particular, it was found that increasing the threshold from 0.1 rho to 0.9 rho was effective in eliminating false alarms due to noise associated with the coaxial connections. Refinement of threshold levels will continue in future applications.

  5. An analysis of the TDR waveforms was a critical component of the plan of action. It was critical to evaluate the reason that the alarm threshold was exceeded and to verify that cable deformation had actually occurred before alerting on-site personnel to conduct a visual check.

  6. The major cause of cable damage was construction operations (i.e., backhoe excavation and drilling of backfill placement holes). The range of movement was comparable with that expected if a sinkhole had developed. Lack of grout did not affect sensitivity which is consistent with performance with similar cable installations on other projects. This emphasizes the need to identify the location of cable runs on the ground surface prior to initiation of vertical drilling operations for mine grouting.

  7. The 3 ft thickness of cementitious backfill placed in the trench installation created a stiff "beam." This thickness should be no greater than 6 inches.


Aston, T., and F. Charette. (1993). Installation and Monitoring of Three Abandoned Mine Crown pillar Sites, Timmons, Ontario: August 1992 to October 1992. CANMET Div. Rep. MRL 93-020(CL), Min. Res. Lab., Ottawa, June, 61 pp.

Bauer, R. A., C. H. Dowding, D. J. Van Roosendaal, B. B. Mehnert, M. B. Su, and K. O'Connor. (1991). Application of Time Domain Reflectometry to Subsidence Monitoring. Office of Surface Mining, Pittsburgh, 48 pp.; NTIS PB 91-228411

Charette, F. (1993). Results of the Monitoring of Three Crown Pillar Sites in Cobalt, Ontario. CANMET Div. Rep. 92-101 (CL), Min. Res. Lab., Ottawa, Mar., 31 pp.

O'Connor, K.M. and C.H. Dowding (1999). GeoMeasurements by Pulsing TDR Cables and Probes. CRC Press, Boca Raton.

O'Connor, K.M. and E.W. Murphy (1997). TDR Monitoring as a Component of Subsidence Risk Assessment Over Abandoned Mines. Int. Journal of Rock Mechanics & Mining Science. Vol. 34,Nos. 3-4, Paper 230.

Willard, D. J. (1995). A Narrow Escape, Tragedy avoided on I-70 despite missed signals. Akron Beacon Journal, August 20, Akron, Ohio.

Forum Home

Updated: 06/27/2017
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000