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

7.4. Phase 3. Observation protocol

7.4.1. Geometry of observations

A guideline issued by Gijima Americas describes the required geometry of observations “it is desirable to keep internal angles as close to 60 degrees as possible for maximum accuracy” [151]. The layout of the mining panels at the South Deep mine are 6 metres wide. In practice sidewall stations are placed 5 metres apart and approximately 20 metres between sidewall station points. In the stopes stations are only installed in the one sidewall as the other sidewall consists of backfilled material that is not considered stable enough for survey control. [150].

The geometry of a resection is described in a Leica report, confirming that “after measuring the second point the location of the total station is ambiguous…” [153], and confirming that “measuring a third point provides redundancy” [153]. From the Kibali Gold Project and PMC Byrnecut Project standards it is noted that the normal mining layout followed in both of these projects is a square pillar type layout which will make the placement of reference stations in excavations joining the main development possible. Such a mining layout would facilitate better geometry and longer distances between the setup position and wall stations. Unfortunately such an ideal layout is arguably not always possible in a narrow tunnel environment.

7.4.2. Maximum and minimum angles and distances

Jaroz and Shepard noted that the shape of a resection will have an influence on the accuracy of the bearing transfer and that a configuration with acute triangle geometry is essential. [96]. In an article published in the Journal of the Institute of Mine Surveyors in 1951, Shewmon observed that “the 2.5 degree, 2.5 degree and 175 degree triangle offers greater accuracy than a 45 degree, 45 degree and 90 degree triangle” [106]. Briggs remarked that provided that the measured angle lies between 165 degrees and 180 degrees and the sides are roughly equal, the angular and linear error will be equal. [154]

The mineMARKUP software user guide indicates that the software has a built-in safety factor that prevents any angle smaller than 20 degrees being observed or stored, which restricts the observer from placing the freestation in a position where this conditions are not met. [151]. The mineMARKUP software of GijimaAST defines an error in a resection setup as a setup where an observed angle is smaller than 20 degrees or larger than 160 degrees, which may possibly provide an inaccurate fix. [155]. According to the standards of Kibali Gold Project, the instrument should be setup in such a position that the observed angles will be “no smaller than 60 degrees and no bigger than 120 degrees” [148] These standards recommend that an ideal base length for extending the survey forward would be 50 metres, stating that “any distance shorter than 25 metres can increase the inaccuracy of the survey, and any error will not show up in the residual results” [148]. The exact same standard is used at PMC Byrnecut Project. McCormack advises that the “longest possible distance between station is practical” and that the best solutions “are obtained by setting up as close as possible to one reference point” [110].

7.4.3. Number of observations

The guideline written by GijimaAST Americas requires that angles are “doubled” when a freestation setup is made implying that the reference object must be observed in both “faces” ⁴¹of the instrument. The standard procedure at Kibali Gold Project calls for a “multiple face” observation. [148]. In this specific case the procedure is made easier with the application of the motorized instrument that will automatically change face and re-orientate the instrument on the target. The “change face” option is not available on the software of non-motorized instruments such as the Leica TPS1200+. As an alternative, on some instruments the “survey forward station” function will allow the surveyor to select a single face or multiple face observations.

4. Phase 4. Calculation methodology

The Leica freestation software makes use of a combination of unique calculations based on observations using unique triangle solutions, resections using the various combinations of observed points and a helmert transformation. An error variance between observed and calculated for each solution from which the frequency of error variance is plotted and an approximation of the freestation point is calculated. The final point co-ordinates are calculated based on a weighting of points and a least squares adjustment made. [111]. If the freestation position is in doubt as a result of poor observations, these observations can be identified and removed or de-weighted using the freestation software. [156]. This would then provide a more accurate final position fix and have the added benefit of identifying points that may have been damaged or surveyed inaccurately.
The mineMARKUP and mineSIGHT software use a combination of all the unique triangles available from the observations to calculate the final position of the freestation. It is obvious that in the case where only two reference points are used, there can only be one solution possible and this could therefore bring the accuracy of the position fix into question. In the case of more than two reference points a number of unique triangles can be solved and the final position determined by the basic averaging of the results. It is unclear if any form of weighting can be included in the solution. In the case where distances as well as angles are measured the solution of the triangle can become difficult due to the number of variables measured. By definition this method is actually a modified triangulation method rather than a resection. This method is used with success at South Deep mines in South Africa with success (please refer to Chapter 6). It is argued that where redundancy is added to the measurements it complicates any attempts at a manual solution of the position fix. The standard procedures used at Kibali Gold Project in the Democratic Republic of the Congo, describes the resection method used is a “form of triangulation using geometric trigonometry software” requiring the positioning of the instrument to ensure “strong geometric features”. [148]

4.