The Surveyors Role In Construction
The Manapouri-Doubtful Sound
Hydro Electric Power Development
R.J. Williams, Dip.Surv., M.N.Z.I.S.
A paper presented to the 86th Annual Conference
New Zealand Institute of Surveyors.
Te Anau, New Zealand – October 1974.
This paper has been prepared as a brief outline of the surveyor’s role in the construction, with particular reference to the Manapouri Underground Powerhouse and Tailrace contracts.
Location and Brief History:
The Manapouri Power Scheme is located in the rugged forested country of Fiordland, approximately 100 miles from Invercargill. It uses the waters of Lakes Manapouri and Te Anau, diverts them through intakes and vertical penstocks to turbines located 700’ underground at West Arm, Lake Manapouri and discharges them through a 6 mile tailrace tunnel to Deep Cove, Doubtful Sound.
A hydro-electric scheme located to use the waters in a 600’ fall through the mountains, was first muted in 1904, but put aside because of the magnitude and inaccessibility of the area.
In the 1940s, a Canadian group examined the power resources of both Manapouri and Te Anau, but no definite proposals were made.
1959 saw the scheme revised when Consolidated Zinc Proprietry Limited of Australia proposed building an aluminium smelter at Bluff, using Bauxite from Queensland and power from Manapouri. The Company’s partnership with the Kaiser Corporation radically accelerated the original construction programme and a separate company known as Comalco was formed to handle the project.
The initial agreement of January 1960, validated by the Manapouri-Te Anau Development Act 1960, granted Comalco, inter alia, the exclusive right to make an economical appraisal of the power potential of Lakes Manapouri and Te Anau, and the exclusive rights to use the waters of lakes and rivers for power purposes to construct and operate generating works. The Company employed the Bechtel Corporation Limited of San Francisco to investigate the site and design the Manapouri Project.
Comalco found itself unable to finance both the power scheme and smelter in addition to its works in Australia and a new agreement was negotiated whereby the Company surrendered its rights to construct and operate the works for its own purposes in exchange for the right to obtain a supply of continuous power from the Crown.
The Manapouri-Te Anau Development Act 1963 authorised the Minister of Electricity to construct and operate the power scheme and to raise and lower the levels of the lakes and rivers.
Preliminary Control Surveys:
Up until the late 1960s Lake Manapouri and its surrounding area was fixed only by a reconnaissance survey made by Jas. McKerrow in about 1863 and by surveys executed by a Me Jenkins in 1926. It wasn’t until early 1960 that a request for triangulation to cover the Lake Manapouri-Doubtful Sound area was made. At that stage an accurate elevation of Lake Manapouri was not available. So with the interest shown in the possible development of the Lakes Manapouri and Te Anau area, the need for accurate survey control for mapping and the determination of accurate lake levels arose.
In early 1960 Lands and Survey Dept personnel levelled across Doubtful Sound to West Arm, Manapouri coming off the benchmark established by Mr Jenkins at Doubtful Sound, mean sea level being established by Mr Jenkins by a series of observations over a period of two weeks. This level run was carried out with two Zeiss NI automatic levels and Flood staves, the mean of four reduced levels being adopted.
About this time an automatic tide gauge was established at Doubtful Sound, visual gauges were set up at West Arm and at Pearl Harbour and levels transferred from one side of the lake to the other. Fundamental benchmarks were established at Manapouri and Te Anau, and a precise level run was done from one to the other.
In 1960 after a preliminary reconnaissance of the Manapouri area to determine the most suitable method of extending the control to Deep Cove, it was felt that the nature of the country was such that a standard traverse would be extremely difficult, and would entail much line cutting through very heavy bush. It was decided that triangulation was possible, but difficulty would be experienced in getting quadrilaterals of good shape.
A tellurometer traverse from the existing control surveys adjacent to the Manapouri township, across to Doubtful Sound and back was adopted as being the most economical and satisfactory control.
Mr Dick Innes, now Chief Surveyor, Christchurch, was put in charge of the control survey and operations started in the summer of early 1961.
In preparing the scheme of control, Mr Innes paid particular attention to ensure that the control over the tunnel area was as rigid as possible. Care was taken to locate two points, each adequately fixed for position, with a good ray to a distant point so that sound bearings could be thrown into the tunnel area.
The very nature of the topography made the designing of a well based and rigid control network extremely difficult, and several vantage points had to be discarded because of the difficulty of access. In some areas it was obvious that even an experienced rock climber would have found access difficult.
At the Deep Cove end, it was found that it was preferable to locate the points close to M.H.W. as the portal site was to be located close to M.H.W. This was to alleviate the possible inaccuracies of observing in a steeply inclined line from the portal station to the other control. The Deep Cove end, being a terribly cramped corner, the lines so located were too short for the tellurometers, and had to be chained across tidal flats.
The angle observations were made with a Wild T.2. – 6 sets with a maximum residual of 3 seconds, but on some high points, because of erratic results, twelve sets were observed and meaned. Considerable trouble was experienced in closing the triangles of the quadrilaterals at the West Arm end. Each angle was observed by 3 different observers and with two different instruments. Each observer did at least 6 sets and, although the individual observers differed by only up to 2 seconds on the mean value of any one angle, all provided triangle miscloses for the four triangles of between 9 to 17 seconds less that 180°00’00”.
At Deep Cove, similar trouble was experienced, where the four triangles closed minus 6”, minus 4”, minus 2” and minus 1”.
A logical explanation for these miscloses was sought, and after carefully examining the results, plumbness of the beacons etc, it was considered that the miscloses were the direct result of sudden changes in the deviation of the vertical throughout the control area. When one looks at the high mountainous slopes on each side of the lake falling steeply down into very deep water, I believe that it is quite conceivable that this is the logical explanation for the miscloses. A similar situation exists at the Deep Cove End.
To accommodate these angular miscloses in the two end quadrilaterals, all the sides and the diagonals were measured by chaining or tellurometer.
The whole tellurometer traverse was computed and the system adjusted by least squares and two sets of coordinates produced – one in terms of the National Grid and the other based on the Mt. York meridional circuit. This information was subsequently conveyed to the consulting engineers – Bechtel Pacific Limited.
It is of interest to note that the Bechtel Corporation, in order to make absolutely certain of the information supplied, arranged for an American survey party equipped with a geodimeter to check on the accuracy of the survey work. This party carried out a single traverse line from West Arm to Deep Cove by modified 2nd Order methods. Ten sets of readings were taken with a Wild T2 equipped with a precise 6.4 second striding level, and all distances were measured with a Geodimeter.
Excellent agreement was obtained with the Lands & Survey traverse and I quote from the “Engineering Surveys” report:
“Using our observed values for the angles and distances, our traverse misclosed against the Lands & Survey Department’s adjusted position of “Cove Pillar” at Deep Cove by an error of slightly over 1 part in 100,000 for position and 1 second for azimuth.
“In appraising the value of this check it should be recognised that while both the Tellurometer, as employed by the Lands & Survey, and the Geodimeter, as employed by Engineering Surveys, are electronic instruments, they function on entirely different principles and are completely unrelated methods of measuring distances.
“In our angular work, it should be noted that our observer Mr Hartley has demonstrated a personal factor for unusual accuracy in many years of experience in Triangulation work. In order to eliminate incidence of error as much as possible in angular work the striding level was used on all angles and sufficient sets were turned to ensure an accuracy better than the prescribed limits.
“Accordingly, it is our opinion that our check of the control surveys on the Manapouri Project demonstrated that the Lands & Survey Department has done an excellent job under difficult conditions. In view of the investment involved in this project, there is no doubt that an independent check was a wise and necessary decision and we now feel that Bechtal Engineers can proceed with full confidence in the accuracy of the horizontal control surveys.”
Both survey parties carried out trigonometrical levelling, the American team closing 0.44 feet high on the Lands & Survey Department’s value of “Cove Pillar”, the American trigonometrical elevations originating at Spey Pillar. This comparison with the Lands & Survey Department’s direct levelling carried out in 1960 was regarded to be satisfactory, considering the large altitude of the traverse stations, and the possibility of considerable deflection of the vertical as evidenced by the horizontal angle observations. Both parties revealed an approximately 10 foot error in the Manapouri benchmark as determined by Jenkins in 1926.
A later 2nd order level traverse was carried out by Bechtel Engineers after the Wilmot Pass road was formed, but details of this traverse are unknown to the writer.
Preliminary Construction Surveys:
Prior to the commencement of construction, many smaller surveys were carried out within the project area – these including stadia traverses of the Mica Burn & Disaster Burn and other streams on the Deep Cove side of the pass. This work was carried out by the Invercargill firm of Moir, New & Jenkins, who also completed a stadia traverse between West Arm and Deep Cove. These surveys were executed to assist in the geological investigation to ascertain the type of rock and faults that could be expected during the tunnelling operations.
Detailed contour surveys of the switchyard, intake, and portal areas were carried out by this same firm to enable detailed design of this area to commence, including the siting of a 700’ pilot shaft by Theiss Bros of Australia. With the information gained in the sinking of the pilot shaft, Bechtel were able to carry on with their design etc, and prepare plans for tender purposes.
For the Wilmot Pass road linking West Arm with Deep Cove, a compass chain and abney traverse was carried out, also by private surveyors. With the “Wanganella” arriving at Deep Cove, construction and clearing was able to commence.
Bechtel Pacific, in designing the whole complex, set up their own grid system, grid North being orientated at 5° off true north. This allowed the downstream direction, i.e. lake to penstocks, turbines to drafttubes, to be called North and the centreline of gateshafts, machine hall and switchboard to run east-west.
A construction grid system such as this has many advantages, especially when most of the buildings and equipment is to be set out along straight lines or at right angles to these lines, as is the case on this project. One is able to see at a glance the relationship between various centrelines when coordinated on this type of grid system.
Having worked on construction projects in New Zealand and Australia where on one project, the centrelines of very large buildings and many miles of conveyors were all orientated at an angle like 51°23’41” to the meridianal grid, orientation to best suit the main centrelines of the various equipment to be installed or buildings to be built.
We have to remember that, on any form of construction job, large or small, it is important that the drawings be easily read and understood, not only by the engineers and surveyors but also by the construction crews themselves.
Construction Alignment and Grade:
On a major engineering project of this nature, close control on construction alignment and grade is of paramount importance. It is essential that all tunnels and shafts be excavated to the specific lines and grades designed, and that the tunnels with opposing headings should meet with sufficient accuracy. This ensures that additional excavation is not necessary to blend in the various headings.
Survey practices are basically similar whether they be on the surface or underground, in both cases the angles are measured by a theodolite, and the distances determined by using a measuring band. Heights are determined with levelling equipment as used on the surface. The difficult of applying these methods to underground surveys has lead to the development of special methods not normally encountered in the day to day surface work. Usually the alignment and level of long tunnels are transferred through shafts, sometimes up to only 12 ft or 20 ft in diameter, or through steeply inclined and spiralling tunnels where the transfer is done with extremely short distances between traverse stations. Thus utmost accuracy is required and recourse made to special procedures.
What must be realised is that many difficulties are met by the underground surveyor whether he is carrying out precise control or less accurate work. Most of the lines between set up stations are usually small due to the atmospheric conditions causing a fog or smog and the instrument misting up.
Another problem is that most curves underground are of a small radius, and the diameter of the tunnel is usually reasonably small. Thus the lines do not have to be very long before they start hitting the rib of the tunnel.
Blackness is a problem, everything is dark as there are few artificial lights about the tunnels. Helmet “cat lamps” and illuminating equipment is a necessity. Dampness is a problem too, and it was found on this project that all instruments, when not in use, or after each shift, had to be kept in a lamp box to dry them out.
As in most hard rock tunnels, the excavation was carried out in a cyclic operation with a sequence of drilling holes into the face of the tunnel heading, filling them with explosives, firing and mucking out the shattered rock. Each cycle was repeated about every 10 feet of tunnelling.
Before drilling, the surveyor was required to “paint the face” – the tunnel profile was spray painted on the heading face, being marked out relative to tunnel centreline, designed invert or springline elevation.
On this project the survey control to the machine hall and beyond was carried down the access tunnel which is 6,700 feet long and spirals at a grade of 1:10 down to the machine hall 700’ below the surface. Invert marks, with offset location points grouted into the rib walls, were used as traverse stations. The horizontal angles were measured using Wild T.2 theodolites. Six sets of angles were observed, the observations being made to Wild Illuminated target equipment. Because of the curves, each line averaged about 200 feet long with the longest line observed being only 800 feet. With the shortness of lines and lack of interchangeable tribracks, extreme care was taken to eliminate plumbing errors. The distances were measured in catenary with a standardised 300 foot steel band, the distances double chained and meaned. The return traverse realised an angular misclose of 17 seconds and a coordinate misclose of 0.041 feet North and 0.014 West. A good result when considering that approximately 50 stations were involved in the total traverse, and is a good illustration of the standard of the work carried out on the project.
The access tunnel control was checked by plumbing down the service shaft using the “Weissbach triangle” method as described by W. Wassermann, M.I.S. Aust. in N.Z. Surveyor, Number 231. This checked within 0.1 ft. and a few seconds for bearing.
The excavation of the machine hall was done from both ends, the survey control at one end coming directly off the access tunnel control. The other was carried down the spiralling emergency exit tunnel and then extended into the machine hall proper. The levels were transferred down the access tunnel by spirit levelling using automatic levels and normal aluminium staves, every reading recorded to 3 decimals of a foot. The majority of the benchmarks consisted of rock bolts grouted into the tunnel ribs. These were found to be reasonably stable in the access tunnel area although evidence of stress relief was apparent in the machine hall and draft tube manifold. In these areas movement of up to 1” was noticed, this being particularly obvious when invert marks were offset for location and redefinition purposes by lead plugs or the like in the tunnel ribs and after a few months it was found that the total distance between the offset marks was somewhat smaller than previously recorded.
As the excavation of the surge chamber lead off the access tunnel and spiralled down into the draft tube manifold, so did the survey control for this excavation and for the West Arm section of the tailrace tunnel. This control was extended through the draft tube manifold area and along a small drift joining this area to the lowest portion of the service shaft and closed by plumbing down the service shaft from the access tunnel control. At the calculated point along the surge chamber a right angle was turned and this led directly into the horizontal curve in the tailrace tunnel. Of the 6 mile 688 feet of tailrace tunnel, 4/5 of a mile was excavated by the West Arm contractor and the remaining 5 1/4 miles excavated and lined by the Deep Cove contractor. Hole through was achieved in October 1968. The mis-alignment between the two headings was approximately 3/4” in the horizontal direction and 4” in the level.
Part of the surveyor’s job during the tunnelling stages was to check the excavation for “tites” and excessive “overbreak”. Various methods were used for this. A lazer beam mounted on a K. & E. transit theodolite was set up on centreline and at the design grade. Offset measurements were taken from the rock wall to the beam and compared with the design dimensions, the tites being pointed up with spray cans at the same time.
Another method used quite extensively on the project was making use of photographic equipment for cross section purposes. Flash units were mounted in a box with a 1/4” slit cut in the sides and top. This box was then mounted in conjunction with a Wild subtence bar over a known survey station and the elevation of the subtence bar recorded. A camera was set up quite a distance back from the flash box. The flash lit up a thin section of both the tunnel walls and ceiling which was recorded on film. These shots were subsequently developed and blown up to a pre-determined scale using the length of the subtence bar as the known distance. The design tunnel profile was plotted on the photograph. Areas of overbreak were measured by planimeter and volumes calculated. These photographs were kept for record purposes and have been used as evidence for quite extensive claims for overbreak and the extra concrete used in the final lining of the various tunnels.
A few problems were experienced in the initial stages of the photographic cross-sectioning, but with check dimensions being done by cloth tapes it was found that the method was quite accurate and was particularly good for the checking of the tites and for record purposes. A black board showing the station number, and elevation of the subtence bar was set up beside the subtence bar so all the information required was recorded on the film.
Once all the tunnelling drew towards completion, the project gradually moved into the construction phase, which involved the surveyors in re-setting out main centreline moments, and re-siting accurate benchmarks. From these points the surveyors set out and levelled all the concrete lifts prior to pouring, and played a full part in the setting of all the steel work in various parts of the project.
To control the concrete lining (lined by slip forming) of the 466 feet penstocks, a lazer beam was used. this was firstly mounted at the lower end of the penstocks pointing vertically upwards but was later transferred to the top of the penstocks and mounted on a frame, rock bolted into the top arch of the penstocks. The beam was projected onto a target located in the lower penstock area. On the slip form platform a small screen was attached. This had a small hole cut out of it to allow the lazer beam to project through and onto the lower target. Whenever the slipform wandered offline a little, the beam would project onto the screen, thus the operator could see at a glance when the form was off line and also the direction in which to correct.
The services of the surveyor were in particular demand in the setting of the scroll cases, dresser couplings and all the gateshaft steelwork, where extremely close setting tolerances were demanded. Seal faces and wheel guide faces and other steelwork in the gateshaft area extending nearly 40’ high were required to be set truely vertical and co planer with ± 1/32”. With extreme care, this was achieved even though the control case line was only 12 foot long.
These are the types of demands to be met by the construction surveyor, and it is only through developing methods and procedures to meet these demands that these limits of accuracy are achieved.
Throughout this paper I have outlined the type of work that is carried out by the surveyor in a large construction project like Manapouri and hope that I have illustrated fully the essential and important role that he plays from the outset and through to the final completion.