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

2.8. The effect of external factors on the accuracy of a survey network

In the underground environment the environmental influences on the accuracy of surveys and the establishment of well-spaced survey stations can be significant and will need to be taken into consideration in conjunction with the limits of error. Young argued in a study that:

“The actual underground conditions frequently necessitate the establishing of stations in the roof instead of in the floor and setting up the transit under a point instead of over a point. Unless protected in some way by a shield or covering, a station in the floor of a roadway along which there is considerable haulage is liable to be knocked out.” [6].

Little has changed regarding the factors that can influence the accuracy of a mine survey in the past hundred years when Young originally made these remarks. In the modern mining environment such factors still include:

• 
  refraction caused by columns of air of different temperatures,
  
• 
  illumination of targets affecting the correct identification and accurate centring of the crosshairs on the target,
  
• 
  the visibility affected by blasting and exhaust fumes as well as water vapour from condensation limiting the distance of clear visibility,
  
• 
  ventilation or air flow affecting the movement of plumb bobs and targets,
  
• 
  heat produced from machinery and in-situ rock temperature,
  
• 
  the height of workings complicating the accurate centring of the instrument ,
  
• 
  the condition of the rock with reference to deformation and scaling and;
  
• 
  the very significant but often overlooked influence of production schedule pressures on the Mine Surveyor.
  

Stengele and Shätti-Stählin lists similar conditions found during the Gottahrd tunnel construction, remarking that “underground measurements always take place under tough conditions,…, permanent time pressure, unfavourable conditions (visibility, light, noise, temperature, humidity, ventilation, traffic...and safety-related constraints…” [59]. These factors will be discussed under individual headings in the following section.

2.8.1. Illumination

The underground environment requires artificial lighting in order for surveyors to accurately identify and observe survey reference stations. The absence of a natural light source can cause complications in the correct identification of points as well as the accurate sighting of targets. “Artificial light is necessary in order to illuminate the point of sight and the cross hairs. Such light is generally very poor, and this fact greatly hampers work with the instrument.” [6]. In a case where multiple targets are observed the incorrect identification and illumination of targets can result in gross errors being introduced into the survey network.

2. The effect of refraction

Böckem , Flach et al, noted that refraction is considered a major source of systematic error in the precise determination of distances and angles [60]. The effect of refraction on the accuracy of a tunnel survey has been studied in the tunnelling environment by Korittke. [61]. Fowler concluded in his dissertation that “Lateral refraction is the biggest source of systematic error that can be encountered in underground surveys. Every effort should be made to avoid its effect. …” [45]

Refraction is caused by a temperature variation between the surrounding air and the ambient rock temperature of the tunnel. If the refraction index is too high, the accurate positioning over targets by the operator could be affected. Schwartz defines the problem as follows: “refraction of air means that if the density of the air is not constant then the line of sight to the target point is not exactly in a straight line. The density depends mainly on the temperature of the air…if there for example is a gradient in temperature of only 0.1 degree/m the maximum offset of the line of sight with a length of 600m is 4.5mm” [62]
In the Gothard Base tunnel the effect of refraction on the azimuth transfer of the survey was studied in detail: “Especially on the occasion of the direction transfer at the tunnel entrance, the optical direction can be extremely affected by refraction effects. There are a lot of concepts to determine the refraction corrections by calculating loop misclosures and using available gyroscopic observations.” [63]. The effect of refraction has been studied extensively in the tunnelling environment where high accuracy surveys are essential to the guidance of Tunnel Boring Machines (TBM’s) and where breakthrough accuracies of less than 10millimeters are required. In the South African mining industry no study has to date been made on the effect of refraction on the accuracy of surveying underground. This is in part as a result of the fact that the method of extending a traverse has been by the installation of pegs in the hangingwall of the excavation. The method of setup required by this method of survey places the line of observation almost in the centre of the excavation, an area where refraction should arguably have the least influence on the accuracy of the survey. It can be argued that the suggested method of installing sidewall survey stations could be influenced by refraction as the targets are now moved from the relative stable column of air in the centre of the excavation to the relatively hot column of air in contact with the rock in the sidewall of the excavation. As the sights would be shorter it could be argued that the effect of error could be a factor in the accuracy of the survey.
Common sources of heat in the mining environment includes the temperature of the air forced underground, compression of this column of air, the distance that the air column must be moved to the workings, rock temperature, mechanical equipment such as ventilation fans, diesel machinery, heat generated by ground water and explosives [64] and the oxidization of timber and minerals. [65].
The geothermal gradient in South African Bushveld complex mines have been calculated to be 21.2° Celsius per 1000 metres, which is considered higher than that of Witwatersrand type mines, so that in Bushveld complex mines. “at depths of 1000 metres virgin rock temperatures of 43 degrees Celsius can be expected” [66]. Comparatively, gold mines in the Witwatersrand complex at the depth of 1000 metres, indicates virgin rock temperatures ranging from 27° Celsius in the Central Rand to 32° Celsius in the Western areas of the complex. Temperatures of 36° Celsius is the norm in mines in the Free State. [65] To compensate for the extreme environment found in these mines, the object of ventilation is to ensure an efficient thermal environment in which air- and heat flow is considered [67] in such a way that “a wet-bulb temperature of 27.5°C is generally maintained” [66]. The Regulations of the MHSA. Chapter 9.2.(2)(b) defines the limits of the environmental control that is required to be established and maintained in all working places by the employer as follows: “ thermal stress – heat >25.0°C wet bulb; and /or >32.0°C dry bulb; and /or >32.0°C mean…” [11]
It was observed by Glaus and Ingesand that the gradient of air temperatures during winter time when cold air is pumped into these tunnels caused up to 10mm differences between lines of sight larger than 100metres [68]. Similarly, Ingesand and Ryf remarked that direction transfers at tunnel entrances are the most suseptible to refraction effects [69]. This effect should be considered when direction transfers are done in incine shafts specifically in winter when a larger temperature gradient may exist between the tunnel and outside temperature. For the purposes of this research the effect of the external temperature gradient will not be considered as the proposed application of the alternative method of surveying will be applied to tunnels away from the shaft entrances. Air must travel vertically down a shaft and then along a horizontal airway while exposed to hot rock before reaching the area which requires ventilation. Dickson and Starfield argued that the heat pick up is small as a result of the similarity between the “autocompression and geothermal temperature gradients” [67]. The authors postulated that an isotherm around the tunnel depends on the wetness of the tunnel. The occurrence of machinery that generates heat such as diesel locomotives may affect local temperatures but is not considered to be significant. [67] Dickson and Starfield argued that the “heat flow into the airways is often controlled by the thermal conductivity of the surrounding rock rather than the surface transfer coefficients from rock to air” [67]
Velasco et al considered that inside a tunnel a “coaxial laminar gas flow at the speed of approximately 2ms^-1 which presumably stabilizes the horizontal and vertical thermal gradients” [70]. The size shape and age of the airway can influence the heat flow within an airway. Whillier states that heat transfer between rock and air never quite reaches a steaty state: “typically after 12 hours about 75 percent of the total temperature drop occurs within the rock…, acceptable accuracy is usually obtainable by assuming the exposed rock-surface temperature to be at the dry-bulb temperature of the ventilation air.” [71], in most cases the dry-bulb temperature of the ventilation flow in a mine is kept below 27° Celsius. It is argued that the rock of the sidewalls of an excavation will assume the average temparature of the ventilated air in that excavation. [72]McPherson clearly illustrates the transient heat conduction equation [73] as follows:

Figure . Variation of rock temperature w.r.t. time and distance

Ramsden noted that “When the rock is exposed the surface rock temperature very quickly reduces to about 2 °C above the prevailing temperature…”, remarking that the main factor influencing the temperature influence is time,, with the temperature stabilizing in approximately two days from initial exposure. According to Ramsden isotherms in tunnels are usually less than 2 °C and usually ignored. [74]
Simpson observed in his text on surface control point surveying that trilateration can be used in conditions where “…lateral refraction has made it impractible to make high quality direction observations with a theodolite” ( [75]. Bannister referred to the effect of refraction when taking observations close to the ground where refraction is the most pronounced, recommending that “… to avoid refraction the lines of sight should be well clear of the ground.” [28] Hirt et al devised an empirical standard deviation form observations made, finding that vertical angles can vary between 1.3 to 3.8 arc seconds [76]. An estimation of the refraction error was proposed by Heister & Liebl from the following equation:

( )
As Chrzanowski (1981) has pointed out, even a very small lateral temperature gradient can produce a severe error if it prevails over the length of a line-of-sight.” [77] .Chrzanowski determined that if the temperature gradient remains constant over the whole observed length, the angle of laterla refraction can be calculated using the following equation [78]:

( )

where:

γ is the lateral refraction,

P is the barometric pressure in millibar,

s the distance of line of sight and;

T is the temperature in Kelvin.
It is argued that in construction tunnels with significantly larger dimensions lateral refraction becomes more noticeable. Advances with a TBM of 20metres per day can be expected. Under these conditions it is possible that the rock temperature has not cooled down as discussed by Le Roux. In addition a larger temperature gradient between the TBM, rock temperature and the ventilation air can be expected. It is argued that this effect should be less in a mine tunnel being developed at a much slower rate. During the survey of the control survey of the superconducting super collider Greening et al observed that “The effect of lateral refraction is well known but remains an intractable problem especially in a tunnel environment.” In an interview Ramsden confirmed these findings by stating that: “When TBMs are operating they produce a large quantity of heat which increases the dry bulb temperature.” [74]
Flach determined that large temperature gradients have the greatests effect closer to the instrument and less critical at the target [79] It is argued that in the case of a freestation type set-up, the instrument will be situated away from the sidewalls of the tunnel and should therefore be affected less by the effects of refraction. Flach determining the effect of refraction on precise levelling observations found deviations of up to 0.54mm over a distance of 60 metres. Refraction can be seen as a significant constraint in the accuracy of high –precision geodetic measurement, however it is argued that the effect of refraction in measurements made for the direction control of a mine tunnel does not have to meet these stringent geodetic accuracy requirements.
Korritke found in studies conducted during surveying of the Gotthard tunnel that “The size of the refraction angle was in the range of up to 2mgon” [61] Heister came to a similar conclusion when he concluded that the effects of lateral refraction cumulative errors is best managed with zigzag traverses supported with gyrotheodolite measurements. [77] Glaus and Ingersand refers to the effects of lateral refaction in of tunnels of diameters of 3-5meters [68] The authors recommends that lines of sight closer than 0.5 metres from the tunnel walls should be avoided as a result of the effect of refraction and recommended “gyro measurements on every fourth traverse leg or every 480 metres” in the CNGS tunnels. In this manner “constant refraction errors” are minimized to 2 mgon. [68]. Flach argued that traditional methods of observation used in geodesy such as the symmetry of observation setups, reciprocal observations and redundant observations “often eliminates refraction influences simultaneously with other systematic deviations.” [79]. Velasco et al advised that in order to control the effects of refraction in a tunnel, “the gyrotheodolite avoid the lateral refraction errors and checks the traverse angular transmission errors… with “cross observations to minimize lateral refraction effect” [70].
Heister advised that making observations from the thermally stable centre line of a tunnel should allow for refraction free measurements, should this not be possible, it is advised that diagonal sightings are “70% less influenced” by refraction [80].
As a result of the extensive work done by various authors, it has been decided that the effect of refraction in the controlled environment of a mine tunnel is to be excluded from this research, an error of 2mgon converts to an error of 0.002/400*360 = 0.0018 degrees = 6 seconds in observation. The method of reducing the effect of refraction by taking diagonal sightings as advised by Heiter combined with reciprical observation proctocol and the use of the gyrotheodolite will be utilized in this research to reduce the error introduced by refraction. There is an opportunity to study air flow and geothermal gradients and the effects thereof in the minig environment at a later stage.

2.8.3. Visibility in the underground environment.

Visibility in the underground environment has been a matter of concern from the pre-electric age where it has been remarked that “…although normal sight conditions usually permit sights around 600 feet, it is advisable to set work points at a distance not more than 300 feet to allow for bad sight conditions due to smoke or fog.” [47]. Visibility in a tunnel is influenced by the humidity within the tunnel, in-situ rock temperature, the strength of ventilation flow, condensation of water as a result of heat generated from the in-situ rock, the operation of machinery, water generated from the drilling operation, the presence of dust particles from blasting and cleaning operations and poor lighting conditions. In addition to the abovementioned issues visibility between survey points can be severely influenced by the presence of ventilation columns, service pipes, mining equipment, cables and in some cases rockfalls.

Due to the cramped conditions experienced within the tunnelling environment such as the conditions described in the Eurotunnel by Korritke “The Service Tunnel, during the construction period, is densely fitted out with the various services required for the driving process, including water and drainage pipes, power supplies, ventilation ducts, signaling and communication cables, and twin rail tracks,..” [81] and the presence of workmen and their equipment such as locomotives or trackless machinery the intervisibility of survey stations can be reduced to a point where the control points becomes useless. The obstructions caused by ventilation pipes and mining services such as water pipes and electricity cables seem to be a timeless and universal problem as similar issues were mentioned by Arthur [43].
Almost one hundred years earlier, Johnston remarked on similar issues around the error in observation caused by poor visibility “the liability of error brought about by the necessity for short and steep sights, the interference of water, bad air, steam, the lack of light and cramped places.” [9]. The unique problems encountered in mine surveying in the underground environment are described in the SME Mining Engineering Handbook “However, underground (surveying) work is quite different: [82]

1. The lines of sight frequently must be carried through constricted openings, often involving short lines of sight and awkward setups.

2. The lighting generally is poor, requiring illumination of the backsights and foresights, and the crosshairs.

3. The ambient conditions often are difficult, including falling water, high temperatures, poor visibility, and heavy traffic through the area being surveyed…,

6. Steep vertical sights often are necessary, requiring the use of special equipment.

7. Many working levels commonly are involved, requiring the transfer of position and orientation to each with a high level of precision.” [82]
Shewmon observed in his paper describing the “Random setup” survey method that “..wall plugs projecting a few inches from the pillar wall can be seen and accurately split for distances more than 200ft in fairly clear air.” [15] However Young stated that “smoky atmosphere greatly reduces the possible length of sight” [6], which observation still holds true today even though the ventilation practices have improved greatly since the 1900’s when this statement was made. The causes of the smoky environment may have changed from being caused by smoke generated by open flame lighting and the poor extraction of blasting fumes to similar conditions caused by the exhaust smoke created by diesel fuelled mining equipment.

2.8.4. The interference of ventilation

Ventilation is to provide fresh air and removing heat and dangerous gasses. [83] In contrast with the problems caused by poor ventilation, strong ventilation flow in primary development ends could be the cause of disturbance when using plumb bobs for alignment and centring of the instrument and targets. The flow of air can increase to a point where it becomes nearly impossible to align an instrument accurately. In order to work in conditions such as these it is often necessary to work on alternative shift times such as weekends or arrange to have the ventilation switched off for a specified period of time. As an expensive alternative an optical or laser plummet can be used to align the instrument, but is generally not used due to the cost of such specialized equipment. In some cases the surveyor has to resort to mechanical means of blocking some of the ventilation by a jacket in front of the instrument during the centring process. The interference of ventilation can become an issue during any check survey procedure. In most cases the original survey station would be installed when the flow of ventilation was not that strong. As the development end is developed further and further away or a holing is effected to improve ventilation flow, the accurate centring under the survey station may no longer be possible, possibly affecting the quality of the final check survey.

2.8.5. The height of workings

Falling from heights have been identified as a contributor to the causes of fatalities in the South African mining industry [84]. As a result, the current drive on health and safety with reference to working-at-heights which is defined as working at a height greater than 1.5 metres, can cause a delay in the work as precautions need to be taken to ensure safe access to the roof by means of ladder, scaffolding or mobile equipment with a hydraulic boom attachment. Rapoport and Heller referred to such difficulties in an article in the Journal of the Institute of Mine Surveyors “…considerable difficulty is usually experienced in reaching the hanging in a concrete lined, inclined shaft. The practice of erecting platforms is cumbersome and laborious...” [54] The delay in production in establishing a roof station under these circumstances can be significant. The height difference between the instrument and survey station can compound the problem of correct alignment and centring of the instrument directly under the survey station due to increased swing of conventional plumb bobs and in the case where laser plummets are used the correct alignment with the exact position of the survey station. Chrzanowski stated that “Centring under roof markers are more difficult than conventional centring above the marked points” [30].

The increased height of excavation leads to difficulty in reading the number of previously installed survey stations, especially after being exposed to the blasting fumes that will quickly oxidise the brass survey station markers. Fly rock from the blasting activities can bend and damage the brass spad and can lead to the incorrect identification of survey station numbers. The SME handbook alludes to this common condition stating that “The points to be measured often are difficult or impossible to reach.” [82].
Instrument alignment error is normally assigned to about 3 millimetres [28] when setting up under a survey station. This accuracy refers to the accuracy of the optical plummet and may not be sufficient in the mining context. Under normal conditions, alignment is done by plumb bob and not optical plummet due to visibility and illumination issues. In cases where the roof height is higher than 3 metres, the distance between the roof and instrument will be larger than the normal instrument height. This problem can be compounded by the presence of ventilation causing bob sway. The verticality of the bob can be affected in string ventilation conditions. In the case where the laser plummet or optical plummet is used the distance compounds the error greatly. These factors may lead to alignment errors and compounded error in distance and direction being carried forward leading to an unacceptable limit of error in the position of survey stations.

2.8.6. Poor ground conditions

Falls of ground¹⁷ and blasting activities could lead to roof survey stations being destroyed or damaged making the relocation of such points extremely difficult. A less frequent problem is that of pegs damaged by improper care by workmen, usually using the survey station as a point from which to attach cables. Gillespie referred to the problem of “mischievous workmen” [16] causing damage to survey stations. Metcalfe summarized the conditions under which surveyors must perform their duties “…the position of the observer and his assistants is an important consideration, for during the operations their minds must be fully concentrated on work in hand and free from anxiety regarding their safety….” [3]

The SME handbook states that “The surrounding rock may be unstable, resulting in movement or loss of the surveying stations, as well as hazard to the surveyors.” [82]. Fowler observed that in tunnelling projects in New Zealand “these surveys will usually restart from the tunnel portal, in order to check for movement of brackets that will quite commonly occur in the tunnel due to geological factors such as convergence or squatting. If the survey does not start from the portal it should overlap the previous survey to check for any movement in brackets.” [45] Young remarked on the suitability of different rock conditions for the installation of survey stations: “In coal mines the roof is generally smooth and the walls rough. In metal mines very often one or both walls are smooth and the roof rough. In tunnels generally the walls and roof are all rough. When walls and roof are equally firm, the smooth surface is to be preferred.” [6]. In deep level mines the movement of the rock strata is a recognized problem but still requires to be quantified.

2.8.7. Moving Machinery

According to the 2013 Mine Health and Safety Inspectorate newsletter for April to June 2010 listed at least 7 fatalities as a result of trackless mobile equipment including incidents where persons where run over or caught between the machine and the sidewall of the excavation. [48] The mine survey network is normally placed in the centre of the hangingwall of an excavation, this necessitates the surveyor to be in the direct path of moving trackbound and trackless mobile equipment. In the case where observations are made the surveyor cannot move the instrument and cannot get out of the way of any equipment. This delays the normal production activities and is bound to lead to confrontation between equipment operators and the survey crew. This issue is closely connected to the next aspect of production pressure.

2.8.8. Production pressure

Since the primary survey network control and the provision of surveying lines for direction and gradient control is normally installed during the normal daily production activities of the mine, it is inevitable that there will be interference between the installation of survey points and the drilling, blasting and cleaning cycle of the mining operations. Under the extreme environmental conditions of heat, noise, poor visibility and risk, the surveyor has to perform his tasks with speed and accuracy, the surveyor is required to plan ahead and foresee problems before they occur, Metcalfe remarks on this , stating that “… In cases where assistants are used the surveyor has to plan the assistant’s actions as well as his own...” [3].

A text from 1904 described this problem, stating that “The survey must be conducted so as not to interfere with mining operations. This requires that the work be conducted rapidly and at the same time accurately” [6]. For that reason it can be argued that “traffic frequently interrupts the surveying party.^” [6]. In today’s mechanized and production orientated environment, this argument probably has even more relevance than in 1904. In a tunnelling handbook by Bichel and Keusel reference to this issue is made and incorporated this factor into the defined minimum standards of accuracy for tunnelling [47].
Regardless of the tunnelling operation, time is money and any delay to the process of drilling and blasting of a tunnel is considered to have a direct impact on the working costs of the operation. Chrzanowski alluded to the impact of production activities on the surveying process, stating that “Survey costs are small in comparison to expenditures involved in tunnel driving. Nevertheless if tunnel driving is held up because of faulty survey work or because of interference of the survey crew with driving operations, resulting losses are large.” The cost analysis of heading costs shown here illustrates the point. Unit bid price of tunnel $3000 per foot, average advance per day 40 feet per day, average income per hour $5000 per hour” [47] Chrzanowski remarked on the cost of tunnelling in 1980 as “it is an expensive operation costing between $10 000 000 and $ 20 000 000 a km depending on the diameter of the tunnel,…” [30]. This information is dated but describes accurately the delay in production caused by a delay due to survey installation of survey control. Such delays must however be weighed up against the costs involved in correcting direction due to misalignment caused by inaccurate survey control.
In underground mines no allowance is made for deterioration of accuracy due to time constraints “The short time available for the performance of these measurements explains the lesser precision required.” [47]. In a paper on the development of new surveying techniques for industry, Dickson related that “Timing is everything …Get to the job too early, and the work can be washed away before the operators can use it. Get there too late and operators are standing down, at a cost...” [85]. Johnson strengthened this argument when he argued that “Too much stress cannot be laid on the importance of the care to be exercised in running connections, as there is nothing the mining surveyor’s reputation depends on more directly than his uniform success in this matter. In fact a failure in such a case may involve a large loss to his employer, or if he has guaranteed the work the cost comes out of his own pocket. On the other hand an error in many cases cannot be remedied, but results in permanent injury to the mine.” [9]. This statement still holds true today as the error in the direction of a mining excavation can cause large financial losses due to the loss in production caused by re-development and non-adherence to tight production schedules and also the financial and legal applications incurred with the loss of life that could result from unplanned breakthroughs into other working excavations or abandoned excavations with an accumulation of gas, water or mud.
The financial implications of an inaccurate survey network that comes under the attention of the DMR may result in the Principal Inspector of Mines being advised to request the mine to perform a check survey to be done on a mine,  as defined in the Mine Health and Safety Act, 1996 (Act 29 of 1996) Chapter 17.11, the cost of such a check survey may be required to be “ …borne by the employer if it is proved that there are errors in any survey or plans constructed therefrom, or that they do not conform to the standards of accuracy required by these regulations” [17].
It could be argued that a check survey may involve substantial financial cost and loss of production. The concept of accuracy and it’s impact on the mining operations had already been described by Agricola in 1556: “Most particular care must be taken that we do not deviate at all from a correct measuring; for if, at the beginning, we are drawn by carelessness into a slight error, this at the end will produce great errors.” [86]. Sir Michael Walker observed at the Conference of Commonwealth survey officers, 1971, Cambridge, England that:

“ The costs of surveying and mapping may well be less than one percent of the total cost of a major project and it seems a relatively small premium to pay to be assured, …that a tunnel started from opposite ends of a hill actually meets in the middle” [87].