Spatial positioning of sidewall stations in a narrow tunnel environment: a safe alternative to traditional mine survey practice

The closed traverse

In the tunnel project a modified “one person” method of surveying was adopted. The targets used consisted of a brass hook with a threaded base, into which standard aluminium target-rods of 300mm long sections were screwed until a desirable length was obtained. A small “grub” screw facilitated the attachment of the rods to the standard mini prism. The target height was measured with a small steel tape as a check on the calculated length determined from the number of rods used at a specific setup. This method of determining target heights provided a more accurate fix of the target lengths than measuring with a tape, as the bending of the tape was avoided. It was observed that great care had to be taken to ensure that all the rods were tightly screwed in, as an unscrewed rod could have resulted in an incorrect target height to be logged. In general, it was found during the tunnel network establishment, that the use of standard rod lengths significantly reduced the time taken to ensure that the target heights were measured accurately. The prism rods also removed the need to access the hangingwall in order to thread a bob-string³² through the survey spad hole. It is possible that the use of standard distances for the target-rods and prism, makes the measurements more accurate than measurements made with a tape that is not always perfectly aligned in the plane of the target-rod.

The levelling and centring of an instrument under a survey station, using modern surveying instruments can be done by using the laser pointer of the instrument as a plummet. In order to align the tripod roughly under the survey station, an assembly of target rods was suspended in a similar manner to that used for the foresight and backsight target establishment. The rods gave a good reference point to align the tripod. The rods were then removed and the instrument placed on the tripod. The instrument was then levelled using the pond-bubble and the telescope swung to point to 0:00:00. With the red laser pointer, the instrument could then be levelled and centred using a similar method to that of the conventional levelling and centring method. It has been observed that in some instances, Mine Surveyors are reverting to measuring the instrument height not by steel tape, but by using the EDM with the instrument telescope set to 0 degrees³³. Such a measurement is not possible without using the reflectorless EDM option. Most modern instruments have this capability. Should the surveyor attempt to make a measurement using the prism mode, the instrument will report an error. Using this method of measuring the instrument height could mean that a surveyor can “forget” to change the EDM setting back to IR (Infra-Red reflector mode) after the measurement has been made. Reflectorless measurement is made to the outside of the spad or any other point that may reflect. Should the surveyor forget to re-set the EDM settings, any further distance measurements will be made using the incorrect prism constant. If the instrument telescope is not set exactly at 0 degrees it may mean that a slope distance, not a true perpendicular distance is measured. In the distance is measured to the peg in the direct “optical centre” of the instrument as compared to the standard method of measuring to the side telescope marking which necessitates the tape to bend slightly. If the method is used throughout the survey the error is constant and can be ignored, however if a change in the method takes place it could mean an error in elevation. The survey network established for the experiment will test this assumption. The accuracy of centring using this method depends on the following factors:

• 
  The quality of tripods used;
  
• 
  Tightness of clamp screw of instrument;
  
• 
  The calibration of the instrument (vertical and horizontal alignment);
  
• 
  Accuracy of levelling of the instrument; and
  
• 
  Accurate determination of the centre of the spad sighted
  

Figure . A spad sighted through the telescope Figure . Laser pointer on spad

Photographs by H Grobler

A calculation of the estimated error in centring was made using the method described by Bannister [28]. The following equation is used to calculate the centring error to be expected at each of the setups. It is evident that the shorter the distance the larger the error to be expected, a fact borne out by the actual survey.

( )

where:

( )


Table . Maximum expected centering error


Due to the short bases encountered it was decided to use forced centring to transfer the survey into the tunnel. A tripod was aligned under the reference point, and aligned and centred using the survey instrument. The alignment was checked using an optical plummet and the conventional plumb bob. Once a satisfactory alignment under the survey point had been achieved, the process was repeated under the survey station and the foresight. Targets were then placed in the levelled and centred target holders under the reference object/backsight and the foresight.

1. Gyroscope Bearing Check

In order to verify the accuracy of the bearing transfer a gyroscope baseline was established on the first leg of the level section of the underground network base UJ022 to UJ004. As a check the bearing was verified from the opposite side of the baseline. The bearing obtained from the gyroscope survey compared well with both the original and check bearing of the baseline.

Table . Gyro baseline calibration.

Gyro Observations	Calculated Bearing
Bearing UJ004 to UJ022	165:15:13
Bearing UJ022 to UJ004	345:16:07
Corrected Bearing UJ004 to UJ022	165:15:40

Figure . Gyrobase in tunnel

The calibration error on the surface baseline was determined to be +0:16:54. Using this information, the underground baseline bearing was determined to be 165:15:40.

The results of the survey are detailed in Appendix 6. Upon completion of the balance of the traverse observations, the corrected bearing of the traverse was compared to the gyroscope bearing that was calculated with the results tabulated in Table . The definition of allowable limit of error on bearing as defined by the MHSA is 2minutes of arc between the original and check survey [11]. The results obtained were within these minimum standards of accuracy, with the original bearing of 165:16:01 compared to the gyro bearing of 165:15:40, an alignment error of 0:00:21. It was decided that no further adjustment was to be made to the observed bearings. The gyro base confirmed the accuracy of the transfer of the bearing through the steep section of the traverse.

2. Calculation of the standard of accuracy for the closure at the breakthrough point

The closure on the breakthrough point before any adjustments were done was determined to be dY = 0.008 dX = 0.003 dZ = 0.000. This closure complies with the Regulations of Chapter 17 of the MHSA Act requirements for the standard of accuracy for a Class “A” survey, determined with equation ( ) which for the distance of 583.371m is 0.034m.

According to the interpretation of the requirements of the regulations, the distance calculated between the closure points was 0.009m, the value of the error vector calculated using the following formula:

( )

which complies with the prescribed standard of accuracy.

Table .Inter station distances and Total length of closed traverse

The calculation of the closure error vector according to the MHSA standards of accuracy is tabulated below:

Table . Standard of Accuracy calculation

The vertical component of the traverse indicated a closure of 0.000m, as a result of which it was decided that a level traverse would not be required to evaluate the vertical error component. The steps followed in the computation of the traverse, was, according to Allan [137] , made in the following manner:

• 
  The closure at the breakthrough point was confirmed to be within the prescribed standard of accuracy outlined by the MHSA to be 0.034m;
  
• 
  The bearing for the first and last line of the traverse on the surface baseline UJ221 to UJ220 was confirmed by the calculation of a Join between these two points;
  
• 
  The traverse field observations were used for the computation of the bearings of the intermediate line;
  
• 
  The sum of the internal angles of the closed figure, consisting of 16 points was determined using the following formula Sum of internal Angles = (n-2)*180°. The error in the sum of the internal angles was determined to be 0:00:07. This error was divided equally between the 16 internal angles observed to obtain the desired sum of the internal angles.
  

The angles within the closed traverse were balanced in the following manner:

• 
  Using the orientation bearing and the adjusted internal angles, the bearings of the intermediate lines were adjusted;
  
• 
  The adjusted bearings were used to calculate Provisional co-ordinates for the network control points. The control points were denoted with a “P” at the end of the number to indicate the Provisional status of the point;
  
• 
  The differences in Y co-ordinates(dY) and X co-ordinates (dX) between the start and end of the traverse were calculated to be dY+ 0.008 and dX +0.003m;
  
• 
  The differences were balanced out over the length of the survey using the Bowditch method using the amount of adjustment as a ratio of the horizontal difference between points; and
  
• 
  Final co-ordinates denoted with the post-fix “F” for final were calculated as the check computation. The differences between the values of the adjusted co-ordinates of the closure point were compared to the co-ordinates of the point of origin was confirmed to be dY 0.000m and dX +0.000m.
  

It was decided to adjust the final co-ordinates of the survey using the Bowditch Rule as this is currently the accepted method of adjustment used on mining operations in South Africa. Although a least squares adjustment of the co-ordinates could be attempted, the basis of this investigation is to replicate current mine survey practice for network establishment and adjustment. Schofield observed that “…the Bowditch rule can be considered the best and…affords a simple, easily applied method of adjustment.” [115]. In research conducted by Schofield it was observed that the method of defining the standards of accuracy has implications in the adjustment method selected. It was observed that the Least Squares adjustment method reflects the angular error in the survey, where in the case of a Bowditch adjustment, this angular error has been adjusted prior to the computation of the error vector. [115]. The final results of the balance of the traverse are listed in Appendix 5.

4.5.3. Possible sources of error in closure

Deakin advised that “good survey practice demands that traverses be closed so that miscloses may be used to assess the precision of the traverse measurements….” [145]. In the underground mining environment such a closure is not possible, unless a holing is affected. Possible sources of error that can occur in any traverse were identified by Allan, Hollwey, & Maynes and include [137]:

• 
  “defective centring of theodolite” under the survey station
  
• 
  “defective centring of the” survey prism or target under the reference survey station. In the underground environment, possible causes could include:
  

• 
  damage to the survey station from blasting
  
• 
  bent or damaged prism target-rods
  
• 
  obstructions under the survey station preventing the target from swinging freely
  
• 
  bent target hooks
  
• 
  strong ventilation currents
  

• 
  “defective levelling of the instrument…” [145]. This error is addressed to some extent by the automatic compensator found in most modern instruments, however it should not be seen as a substitute for good survey practice. In a working end in the underground environment there is always the possibility of the instrument being bumped by passing workers and mechanized equipment or as a result of slippage caused by poor footwall conditions.
  

• 
  “the theodolite not adequately secured ..” to the tripod by means of the clamping screw. If the instrument is not clamped securely, there is a good chance that the instrument will shift during vigorous plunging and turning of the instrument during observations. In the case of automated and motorized instruments this point becomes even more significant.
  
• 
  Allan, Hollwey, & Maynes refer to the “shifting head” and “lower plate” being unclamped [137]. Although this problem may not occur with modern instruments, it is possible for the surveyor to forget to “set” the reference bearing properly on the instrument, leading to incorrect observations being made.
  
• 
  In addition some instruments have the ability to read clockwise or anti-clockwise readings, and an inexperienced surveyor may inadvertently set the instrument to read anti-clockwise observations, which may in turn lead to great confusion when bookings are made manually
  
• 
  
  

• 
  “Residual parallax “with different positioning of the eye, different objects can be bisected with the crosshairs [137]” In the case of steep and short sights such as those experienced at the transfer points to the tunnel this point became very important. Large errors in bearing were observed as a result of the extreme sightlines that had that had to be used. In addition the effect of refraction as a result of the temperature differential between the outside and inside temperatures of the tunnel in the case of decline surveys. In the poorly lit conditions found in tunnels it is possible that the surveyor can make incorrect sightings as a result.
  

During this phase of the investigation, it was found that the markings on the outer frame of the large prisms used did not necessarily align with the optical centre of the prism, specifically if the prism had to be tilted to be observed in the steep areas of the network. At close distances, using the outside alignment marks on a prism could have contributed to the error in both the horizontal and vertical angles observed.

It proved difficult to sufficiently illuminate the prisms even though ambient light in the form of fluorescent lights were available throughout the entire tunnel. A spotlight was obtained to facilitate better illumination of the targets. As a further refinement, the red-laser pointer of the instrument was used to illuminate and precisely centre all observations made to the reference objects. This method proved very effective but reduced the battery life of the instrument and increased the amount of eye-strain encountered by the observer.

4. Final co-ordinates for the hangingwall control network

The final co-ordinates determined for the survey network to be used for the establishment of sidewall survey stations are listed in Table . The postfix “F” indicates that the final co-ordinate of the point has been determined.

Table . Co-ordinate list of Final control points.

5.