AB
Guidance on Subsea Metrology
International Marine Contractors Association IMCA S 019 www.imca-int.com February 2012 AB
The International Marine Contractors Association (IMCA) is the international trade association representing offshore, marine and underwater engineering companies.
IMCA promotes improvements in quality, health, safety, environmental and technical standards through the publication of information notes, codes of practice and by other appropriate means.
Members are self-regulating through the adoption of IMCA guidelines as appropriate. They commit to act as responsible members by following relevant guidelines and being willing to be audited against compliance with them by their clients.
There are two core activities that relate to all members: u Competence & Training u Safety, Environment & Legislation
The Association is organised through four distinct divisions, each covering a specific area of members’ interests: Diving, Marine, Offshore Survey, Remote Systems & ROV.
There are also five regional sections which facilitate work on issues affecting members in their local geographic area – Asia -Pacific, Central & North America, Europe & Africa, Middle East & India and South America.
IMCA S 019
This guidance has been produced by IMCA, under the direction of the Offshore Survey Division, by Simon Barrett and Jose M Puig of DOF Subsea UK, with the assistance of Keith Vickery of ZUPT and Frank Pritz of Parker Maritime ASA.
This guidance has undergone technical review by members of the the OGP (the International Association of Oil & Gas Producers) Geomatics Committee.
www.imca-int.com /survey
The information contained herein is given for guidance only and endeavours to reflect best industry practice. For the avoidance of doubt no legal liability shall attach to any guidance and/or recommendation and/or statement herein contained.
© 2012 – International Marine Contractors Association Guidance on Subsea Metrology
IMCA S 019 – February 2012 1 Executive Summary ...... 1 2 Glossary ...... 3 3 Introduction ...... 5 3.1 Terminology ...... 6 4 Subsea Metrology Requirements ...... 7 4.1 Typical Required Accuracies ...... 9 5 Subsea Metrology Survey Methods ...... 11 5.1 LBL Acoustic Metrology ...... 11 5.2 Diver Taut Wire Metrology ...... 15 5.3 Digital Taut Wire Metrology ...... 16 5.4 Photogrammetry ...... 17 5.5 INS Metrology ...... 20 5.6 Subsea Metrology Systems Compared ...... 22 6 Subsea Metrology Deliverables ...... 23 6.1 Computations ...... 23 6.2 Reporting and Documentation ...... 24 7 References and Further Reading ...... 25
Appendices A Dimensional Control Requirements for Metrology ...... 27 A1 Rotation of Dimensional Control Offsets ...... 27 B Typical Subsea Metrology Diagram ...... 29 C Comparison of Subsea Metrology Systems ...... 31 1
Executive Summary
Subsea metrology is the process of acquiring accurate and traceable dimensional measurements for the design of subsea structures, primarily interconnecting pipelines. Pipeline interconnections are required to join subsea assets to complete the flow of hydrocarbons from the reservoir to processing and storage facilities. The objective of the subsea metrology survey is to determine accurately the relative horizontal and vertical distance between subsea assets, as well as their relative heading and attitude. This information is then used by pipeline engineers to design connecting pieces to join the assets together. This document explores five of the most common subsea metrology techniques related to interconnecting pipelines – long baseline (LBL) acoustics, diver taut wire, digital taut wire, photogrammetry and inertial navigation systems (INS).
The purpose of the document is to: u Describe, compare and contrast the techniques; u Provide information on the techniques which may be useful to surveyors and surveying organisations; vessel personnel (marine, diving, ROV, etc.); design engineers, fabricators and client organisations. An example is the accuracy of the metrology and how this impacts fabrication tolerances/fit of the interconnecting pipeline, where lack of fit influences the ability to install and create a pressure tight joint as well as possibly influencing the working life of the interconnection.
Long baseline (LBL) acoustics is the most commonly used subsea metrology technique in use today. This method is most widely used because it is adaptable, has redundancy and the results can be processed within hours. It is also attractive because the results can be referenced to an absolute datum. The disadvantages are that it is susceptible to subsea noise and it is equipment and time intensive.
Diver taut wire metrology is essentially a tape measurement of the direct distance between hubs. This method was the first subsea metrology procedure employed by divers and was designed primarily for diver operations on horizontal spools. It is still widely used.
Digital taut wire is a more sophisticated version of the diver’s tape measurements. Additional sensors provide a more accurate distance measurement; depth is also resolved with pressure sensors and relative hub attitude with digital inclinometers. However it still requires line of sight and is not redundant. There is a limitation on the length of spool measured once the weight of the wire causes sagging giving a linear distance error.
Photogrammetric survey has only recently been developed successfully for subsea metrology applications. The basis of photogrammetry is to build a three-dimensional model based on a sequence of two-dimensional photographs. Measuring bars placed on the seabed and reflective markers on the structures provide scaling and reference. The processed images are used to derive a three-dimensional model of the positions of the hubs, the seabed and any other points of interest on the subsea structures. The main advantage of this system is that in a single survey a very high quantity of information can be gathered. The image processing required makes very intensive demands on computer time. Photogrammetry requires good subsea visibility.
IMCA S 019 1 INS metrology is relatively new to the offshore industry. The use and availability of inertial navigation systems has greatly increased in recent years. Inertial navigation systems (INS) use three accelerometers and three gyros to compute a position based on a known start point and the measured changes in velocity and attitude. Unaided INS do not need an outside signal or reference to compute a position; because they are self-contained they do not require line of sight, nor are they affected by poor subsea visibility or a noisy subsea acoustic environment. The main drawback of INS metrology is that inertial sensors have drift associated to them. This sensor drift increases over time and requires some form of correction, generally provided by input from other positioning systems.
2 IMCA S 019 2
Glossary
CAD Computer aided design
C-O Computed minus observed correction
CTD Conductivity, temperature and depth sensor
DSP Digital signal processing
DVL Doppler velocity log
DWG Drawing
DWPLEM Deep water PLEM
EDM Electronic distance measurement
FOG Fibre-optic gyro
IEEE Institute of Electrical & Electronic Engineers (UK)
IMCA International Marine Contractors Association
IMU Inertial measurement unit
INS Inertial navigation system
ISA Inertial sensor assembly
LBL Long baseline
OP Observation point
PLEM Pipeline end manifold
PLET Pipeline end termination
QC Quality control
RLG Ring laser gyro
RMS Root mean square
IMCA S 019 3 ROV Remotely operated vehicle
SVP Sound velocity profile
TRF Terrestrial reference frame
UTM Universal Transverse Mercator
USBL Ultra short baseline
UNESCO United Nations Educational, Scientific, and Cultural Organization
σ Sigma. Represents one standard deviation from the mean
4 IMCA S 019 3
Introduction
This document provides guidance on the most commonly used subsea metrology techniques in use today. These are long baseline (LBL) acoustics, both diver taut wire and digital taut wire, photogrammetry and inertial navigation systems (INS). It covers the basics of subsea metrology, engineering requirements, the different methods and technologies, and some of the advantages and limitations of each technique. The document does not compare or evaluate different manufacturers’ products or services, or the specific performance of systems, and does not endorse or recommend a specific type, model or make of system. However, it should be noted that the use of diagrams and references to proprietary elements and systems may be necessary in a specialised technology such as acoustic positioning.
LBL acoustic systems and techniques are covered in more detail as these systems are generally the most adaptable and most widely used in the industry today. However, it should be noted that the pace of technical change and on-going development of deep water fields means that other subsea metrology methods are being developed, including photogrammetric metrology and INS metrology.
The objective of subsea metrology is to determine accurately the relative horizontal and vertical distance between subsea assets, as well as their relative heading and attitude. Most commonly this is for pipeline connections and the document uses this work as an example throughout. The information determined by subsea metrology is then used by pipeline engineers to design a connecting piece to join the assets together. These connecting pieces are fabricated from steel allowing for some flexibility; however they require tight fabrication tolerances to ensure they meet their intended design life. Their design is twofold. The primary aim is to connect the pipeline terminations; however pipeline sections stretch and contract due to changes in temperature and pressure of the hydrocarbon products being conveyed. The stresses of these movements focus on the interconnection points, so flexibility is also built-in, by designing connecting pieces with one or more bends. Construction tolerances vary with each subsea design but normally are in the order of a decimetre for hub positions and one degree of arc in attitude.
It is often the case that the connecting pieces are the last sections of the pipeline to be fitted and one of the final steps before first hydrocarbon production. For this reason, it is important that subsea metrology surveys are carried out in a timely and accurate manner. If the connecting pieces are not to the required specification and/or do not fit correctly, they can have a significantly reduced life span or can cost days of a construction vessel’s time to repair.
Historically, the first subsea metrology procedure employed was a diver with a tape measure working from flange to flange. However, increasing requirements for greater accuracy and the tighter construction tolerances for deep water field developments, combined with the limitations on depth experienced by divers, have brought about alternative and higher accuracy subsea metrology methods.
IMCA S 019 5 3.1 Terminology
Some of the most common terminology associated with subsea metrology is outlined here. Terminology may be subject to change owing to the swift development of technology and practice in the subsea engineering industry. Subsea infrastructure designs, tolerances, installation and survey requirements, even nomenclature, all are subject to change. u Hubs refer to the ends of interconnecting pipelines which are joined to the subsea assets by hub connectors. Hubs are connectors that are closed together and sealed using external hydraulic pressure rams; these are of modern design and were developed for deep water ROV-aided installation. For the purposes of this document we will refer to all connectors as hubs; u Flange is the term sometimes used for a hub connection that is achieved through bolting together the hubs. These are older solutions developed for diver-aided installation.
The vertical separation of the hubs defines two widely used terms for pipeline interconnections. If much of the length of interconnecting pipe runs horizontally along the seabed it is called a spool , if the pipe inflection design is vertical it is called a jumper . u Spools are normally a shallow water design that allows for protection covers to be fitted to subsea assets to guard against fishing activities like trawling. They normally rest on the seabed; u Jumpers are used in deep water fields where protection covers are not needed. The jumper is designed in the vertical and normally does not rest on the seabed, hence the name, because the pipe piece appears to jump from one structure to the other.
Figure 1 – Example of vertical jumpers
Figure 2 – Example of a horizontal spool
6 IMCA S 019 4
Subsea Metrology Requirements
Subsea metrology surveys are used to determine the relative three-dimensional position and attitude of the hubs and the depth of the seabed relative to the hubs, along with a spool route profile or bathymetric information. Absolute positioning of the hubs is not necessary because it is only necessary to know the three-dimensional range and bearing from one hub to another.
A primary issue is establishing from where on the structure measurement should start. Ideally this should be as close as possible to the hub centre. However, this is not always practical or even possible. The hub might have a pressure cap, the instrument package might be too big to fit on to the hub or access to the hub might be restricted by the frame of the structure, etc. Therefore an offset sensor mount is created, called the observation point (OP). It should be noted that photogrammetric techniques do not require a sensor mount on the structure; only reflectors are placed on the structures, unless the hubs have restricted access and are not accessible by ROV. The actual measurement instrument (the camera system) is mounted on the ROV. For all other metrology techniques an offset observation point is generally required. There are many different solutions for mounting sensors, depending on the instrument – how much it weighs, what measurement procedure is required, etc. For many subsea applications the most widely used solution is a female receptacle on the structure and the instrument mounted on a male stab.
It is normal practice to have one of the hubs in the pair as the spool datum; hub differences and angles are computed relative to this datum hub. There are many criteria for spool reporting; the most widely used is direction of flow. The hub that is first in the direction of flow of the pipeline is selected as the datum hub. This criterion changes from project to project and is sometimes a subjective selection.
Care should be taken when selecting an offset observation point. There are two main priorities: u minimising the offset distance; u accessibility for ROV or diver.
A precise and known relationship between metrology sensors and hub reference points is critical to the metrology process. This requires both appropriate mechanical interfaces and specific high accuracy onshore measurements to enable the computation of metrology values from sensor observations. Dimensional control surveys are discussed in Appendix A.
As far as possible, interfaces for sensors should be pre-built into the structure rather than retro-fitted once the structure is subsea. Dimensional control offsets can only be applied accurately if structure heading, pitch and roll is determined accurately once installed subsea. This needs to be considered in the error budget.
IMCA S 019 7 The inaccuracy of the heading, pitch and roll measurement results in inaccuracy when calculating the observation point to hub offset in the terrestrial reference frame (TRF). This effect behaves like a lever arm; the bigger the offset the bigger the inaccuracy. This is why placing the instrument mount as close as possible to the hub is important. This effect should always be considered when calculating error budgets.
Figure 3 – The lever arm effect of angular accuracy on estimation of position
The instrument mounting arrangement should be as robust and rigid as possible. Each method requires a different mounting mechanism that is mostly tailor-made for each metrology project. Some examples of instrument mounting solutions are presented here.
Figure 4 – The metrology observation point should be mounted as close as possible to the hub
Figure 5 – Examples of male metrology stabs and female receptacle
8 IMCA S 019 As discussed, there are two types of spool design defined by their vertical design. In general, horizontal spools are connected with horizontal hubs; vertical spools are connected to upright or vertical hubs. However, the different hub arrangements have distinct survey requirements associated with the engineering design of the spool.
In general a metrology survey should determine: u the hub to hub true slant range; u the hub to hub depth difference; u the attitude of each hub (heading, pitch and roll); u spool azimuth, relative or absolute; u spool approach angles; u altitude of the hubs above seabed; u vertical profile along spool/jumper route.
And, depending on the hub verticality, the survey should determine: u horizontal hubs: – if the hub is self-rotating or does not have a roll datum, then only heading and pitch of the hubs are required – the approach angles of the spool relative to the perpendicular hub faces are critical. They influence the structure hub to spool hub alignment. Poor alignment will result in a poor or failed hub to hub seal or even in some cases a complete spool misfit; u vertical hubs: – the flat horizontal face of the hub requires determination of pitch and roll, however unless the spool has some sort of key or foot rest, heading has no meaning for the hub face – because the hub heading has no reference, resolution of the spool approach angles to the hub face requires less accuracy than for horizontal spools.
4.1 Typical Required Accuracies
The error budget for metrology depends on permissible hub misalignment defined by spool stress analysis, connector make-up capabilities and spool fabrication tolerance. The accuracy required depends on the subsea connector technology, however nominal accuracies can be stated as:
Point X Y Z Pitch Roll Heading Unit mm mm mm Degrees Degrees Degrees Hub to hub relative distances 50 to 150 50 to 150 50 to 150 Hub to hub relative angles 0.5 to 1.0 0.5 to 1.0 0.5 to 2.0
Table 1 – Typical required accuracies
IMCA S 019 9 10 IMCA S 019 5
Subsea Metrology Survey Methods
5.1 LBL Acoustic Metrology
Acoustic metrology is the most commonly used technique in use today. It is a flexible technique; the equipment is extensively available, supported by the majority of offshore survey contractors and is not solely used for metrology. Long baseline (LBL) techniques are employed to provide an accurate hub to hub range. A pressure/depth survey then determines the hub depths, and subsea gyros and instrumented transponders are used to measure the hub pair’s attitudes. An accurate determination of the speed of sound in seawater is essential to the accuracy of this metrology method. Direct measurement is favoured using a real time sound velocity probe. If calculating the speed of sound from conductivity, temperature and depth sensor (CTD) measurements, care should be taken to use the correct speed of sound equation for the working depth.
Figure 6 – Hub to hub range
This method is most widely used because it is adaptable, has redundancy and the results can be processed within hours. Arrays can be pre-planned to encompass multiple metrologies and seabed structures. It is also attractive because the results can be referenced to an absolute datum. Another advantage is the equipment may already be in use for structure installation so a separate mob of equipment and personnel may not be necessary. The disadvantages are that it is susceptible to subsea noise and it is equipment and time intensive.
Further information on LBL techniques for deep water positioning, including an appendix on the speed of sound in water, can be found in IMCA S 013 – Deepwater acoustic positioning .
IMCA S 019 11 5.1.1 LBL Array Design
The main consideration for array design is that the baselines are of similar lengths and similar acoustic travel time, so that the scaling error in the uncertainty in speed of sound measurement then results in similar baseline inaccuracy. In other words, the longer the two-way travel time the greater the effect of inaccurate speed of sound. Thus unnecessarily long baselines introduce more error to distribute in the least squares solution. A geometric shape with sides of lengths similar to the hub to hub baseline is desirable. A minimum of five transponders is necessary to provide sufficient redundancy to mathematically detect an erroneous baseline in the array. A four transponder array, referred to as a braced quadrilateral , does not have enough observational redundancy to determine mathematically which baseline is erroneous in the array.
A typical metrology array is then composed of two transponders, one at each of the hubs, and three seabed transponders. The seabed transponders should ideally be placed in suitable stands to provide sufficient height and immobility for optimal line of sight. One of the seabed transponders can even be placed approximately 10m along the pipeline to provide an acoustic pipeline or structural heading. A known baseline length on a structure can also provide an independent acoustically derived estimate of the speed of sound in seawater. An acoustic network solution of this design provides a good ratio of observational redundancy to cost.
Sometimes the braced quadrilateral is cost-effective especially in places with good hub to hub line of sight. The need to determine accurately which baseline is erroneous by using a five transponder array might be more time consuming than using a braced quadrilateral. In the case of a bad baseline, sometimes reconfiguring the array and re-measuring affected baselines might be quicker. Having more transponders in an array means more baselines to measure and especially more transponder depths to determine in the depth loops.
Figure 7 – Typical array design for LBL metrology
12 IMCA S 019 Figure 8 – The braced quadrilateral
Figure 9 – Seabed tripods
5.1.2 Stages in an LBL Metrology Survey
The main stages of an LBL metrology survey are planning, preparation, execution and reporting:
u planning – The overall methodology is designed and proposed to the client. This methodology should consider all possible error sources and contain an error propagation budget, and also take account of any schedule and resource constraints;
u preparation: – all instrument mounting hardware and seabed transponder tripods are sourced/fabricated – after all hardware has been installed on structures to construct or define the metrology observation point then a dimensional control survey should be carried out to determine the relationship of this point to the hub and to any other parts of the structure – calibration of instruments and offset determination – prior to the offshore phase of the operation all subsea gyros must be calibrated for relative and absolute C-O values
IMCA S 019 13 – offsets for transponder heights and quartz pressure sensor heights must also be determined – subsea trials as necessary – sometimes it is advantageous to test the docking system for the instruments in a controlled environment;
u execution;
u reporting and quality control.
5.1.3 Equipment List
Using the example of a five transponder metrology array using closed loop pressure surveys, a typical equipment list may be as follows (no spares are considered):
u three seabed transponders;
u two inclinometer transponders;
u quartz or piezoresistive pressure sensor;
u CTD probe;
u direct read sound velocity sensor;
u ROV acoustic transceiver;
u surface command unit and processing software;
u metrology tooling (stabs, handles, frames, work basket, etc.);
u one subsea gyro;
u online survey computer;
u offline computer with CAD package.
5.1.4 Acoustic Metrology Computation
The elements defined in a metrology computation are the following:
u depth determination of observation point from the depth loops;
u least squares adjustment of acoustic array holding depth to pressure transducer observational accuracy (normally 5cm), the same as the baseline weights;
u process attitude data per metrology observation point and compute average heading, pitch and roll for each structure;
u compute observation point co-ordinates from associated transponder co-ordinates, observed structure attitude and transponder height offset;
u compute hub co-ordinates;
u compute spool dimensions.
5.1.5 System Accuracy
Nominal accuracies (excluding error related to offset computations between sensors and hub reference points) can be stated as:
Measurement Accuracy Distance Current DSP LBL acoustic systems offer accuracies of better than 5cm Depth Quartz pressure sensors can support highly precise relative depth measurements to provide an uncertainty of approximately 5cm in 1000m water depth Attitude Manufacturer dependent, but better than required accuracies for metrology
Table 2 – Acoustic metrology – system accuracy
14 IMCA S 019 5.2 Diver Taut Wire Metrology
Taut wire metrology is essentially a tape measurement of the direct distance between hubs. Often the taut wire mounting system is offset from the hubs, and so will require additional tape offset measurements from the wire datum to the hub datum.
The metrology system consists of two ‘jig’ plates with protractor markings etched on them, mounted directly above each of the hubs in a stab-receptacle assembly or bolted onto one of the flange bolts. One jig plate is the anchor and the other is the reel jig. The plates are used to measure the wire departure angle relative to hub headings.
The reel or winch has a device that can measure how much cable has been paid out, or the wire itself is marked off. The jigs have to be levelled and aligned with their respective hub headings; the diver offset measurement also has to be aligned with the vertical and horizontal as much as possible. The wire is paid out, anchored and then tensioned by a hand cranked winch. The readings of distance and departure angle are then observed by the diver. This method was the first subsea metrology procedure employed by divers and was designed primarily for diver operations on horizontal spools. It is still widely used.
Additional measurements may include: u depth survey of the hubs using appropriate pressure sensors; u diver hand-held inclinometer measurements of the inclination of the hub faces; u absolute heading and pitch of the hubs using subsea gyros mounted at the observation point; u locking the wire, recovering the wire to deck and taking an independent measurement to confirm the distance.
The biggest drawbacks of the diver taut wire method are: u it requires direct line of sight (uneven seabed should be evaluated prior to operations); u it has little redundancy; u it cannot be used in deep water and needs good visibility; u readings depend on the observational abilities and consistency of the divers (though potential errors can be mitigated by using different divers to read off the angle and distance values).
Figure 10 – Examples of a diver taut wire system with jig plates graduated similar to a protractor
The accuracy of the diver taut wire method depends on the correct alignment of the jigs and the accuracy with which the length of taut wire deployed can be measured. Sagging of the taut wire will increase with length. This degrades the accuracy of direct distance measurement. Further error sources are the elasticity of the wire under tension, and contraction of the wire due to temperature changes.
5.2.1 Equipment List
u taut wire anchor jig;
u taut wire reel jig;
u taut wire jig to hub adaptor plates;
IMCA S 019 15 u spirit level;
u tape measure;
u folding rule;
u installation tools such as torque wrenches, bolts, cargo strops etc.;
u additional measuring equipment such as quartz or piezoresistive pressure sensor system, hand-held inclinometer, etc.
5.2.2 Method Accuracy
It is very difficult to generalise the accuracy of the taut wire system. Generally such systems are sufficiently accurate for spool lengths less than 10m for a straight spool, but have been successfully used for spool lengths in excess of 30m.
5.3 Digital Taut Wire Metrology
The digital taut wire method is a more sophisticated version of the diver’s tape measurements. Additional sensors provide a more accurate distance measurement; depth is also resolved with pressure sensors and relative hub attitude with digital inclinometers. However it still requires line of sight and is not redundant. It has been primarily developed for ROV operations, but can be diver operated. The tension of the wire is measured digitally and is calibrated before each deployment. The system can also measure vertical and horizontal wire departure angles, and the inclination of the hub is measured with digital inclinometers inside the sensor package. The system has the same anchor-reel principle as does the diver taut wire technique; however the system needs to be powered via the ROV or a dedicated umbilical.
Figure 11 – Side view of digital taut wire measurements and related metrology computations
16 IMCA S 019 Figure 12 – Top view of digital taut wire measurements and related metrology computations
The digital taut wire method can also be augmented with pressure sensor measurements and gyro observations of hub attitude. The digital taut wire metrology method can resolve: u slant range; u horizontal distance; u hub height difference; u pitch of the hubs.
5.3.1 Equipment List
u two sets of metrology docking systems, pre-installed at each end of the spool;
u two digital taut wire measuring units;
u one online computer;
u one offline computer.
5.3.2 Accuracy of the System
Nominal accuracies (excluding error related to offset computations between sensors and hub reference points) for currently available systems are:
Measurement Accuracy (up to 100m spool) Resolution Taut wire distance Error of 1mm per metre measured 0.002m Azimuth angle ±0.5º ±0.1º Elevation difference ±0.03m ±0.01m Pitch and roll ±0.25º ±0.1º
Table 3 – Digital taut wire systems – typical accuracy
5.4 Photogrammetry
Photogrammetric survey methods have been around for some time, but have only recently been developed for use in subsea metrology. The basis of photogrammetry is to use triangulation to build a three-dimensional model based on a sequence of two-dimensional pictures. By taking photographs from at least two different locations, so-called ‘lines of sight’ can be developed from each camera to points on the object. These lines of sight or rays are mathematically intersected to produce the three-dimensional co-ordinates of the points of interest.
IMCA S 019 17 A specialised multi-camera system is deployed on an ROV and sequences of photographs are taken along the intended spool route. Measuring bars placed on the seabed and reflective markers on the structures provide scaling and allow references in the picture sequence. The images are processed using software to derive a three- dimensional model of the positions of the hubs, the seabed and other points of interest on the subsea structures.
The main advantage of these systems is the potential high accuracy of the results. The disadvantages are the very intensive demands on computer time, the requirement for good visibility and specialist personnel and equipment.
5.4.1 Photogrammetric Metrology Computation
Photography represents the real three-dimensional world in two-dimensional images. Photogrammetry aims to reverse the photographic process and reconstruct a three-dimensional model from two -dimensional images. Some information is lost in the photographic process, primarily the depth. For this reason, the three-dimensional world cannot be reconstructed completely from just one photograph. As a theoretical minimum, two different photographs are required to reconstruct the three-dimensional world, and in practice, the solution is to take many more photographs and use the extra information in them to improve the process. The end result of photogrammetry is a dataset of three-dimensional co -ordinates produced from measurements made on multiple photographs.
The principle of triangulation is used to produce three-dimensional point measurements. By mathematically intersecting converging lines in space, the precise location of any given point can be determined. In the case of theodolites, two angles are measured to generate a line from each theodolite. However, photogrammetry can measure multiple points at a time with virtually no limit on the number of simultaneously triangulated points. In the case of photogrammetry, it is the two -dimensional (X and Y) location of a target on the image that is measured to produce a line. By taking photographs from at least two different locations and measuring the same target in each photograph a line of sight is developed from each camera location to the target. If the camera position and aiming angles (together called the orientation) are known, the lines can be mathematically intersected to produce the XYZ co-ordinates of each targeted point. Obtaining the camera position and aiming angles is a process referred to as resection .
The process of resection uses previously surveyed scale bars or known coded targets on the subsea structures. Photogrammetric metrology is therefore a technique that relies heavily on dimensional control. Resection requires knowledge of both the position of the camera and also the direction in which it is aimed. Therefore six values are required to define any given photograph – three co-ordinates for position and three angles for the aiming direction. Additionally the resection can be strengthened by the addition of attitude and pressure sensor information at the camera location. A good resection requires at least twelve well-distributed points in each photograph. If the XYZ co-ordinates of the points on the object are known, the orientation of the camera can then be derived.
The cameras need to be precisely calibrated to remove errors. The triangulation of the measured points is then solved iteratively using the least squares technique using known points like scale bars, control points, the camera calibration values and, very importantly, a ‘first guess’ orientation for each photograph.
At the end of the photogrammetry process a three-dimensional model is constructed that has:
u XYZ co-ordinates (and accuracy estimates) for each point;
u XYZ co-ordinates and three aiming angles (and accuracy estimates) for each picture.
The accuracy of a photogrammetric measurement depends on several inter-related factors. The most important of these factors are:
u the resolution and quality of the camera;
u visibility;
u the size of the object being measured;
u the number of photographs taken;
u the geometric layout of the pictures relative to the object and to one another.
18 IMCA S 019 5.4.2 Equipment List
Marking equipment consists of single markers and scale bars. The scale bars may be single or connected into frames, and may vary in length up to several metres. Single markers can be used separately or connected to frames to enable many markers to be fitted to structures at once. Markers may be installed and surveyed onshore to save vessel time offshore.
Flexible markers (and single markers) can be used to determine cylindrical diameter, ovality and centreline.
A photogrammetric survey spread generally consists of:
u camera system with appropriate lighting and flash operated by diver or ROV;
u deployment basket with metrology marking equipment;
u inclinometer;
u depth sensors;
u processing PC and software.
Figure 13 – Subsea structure marked onshore before installation
Figure 14 – Flexible markers installed onto subsea structure
5.4.3 System Accuracy
The accuracy for any point measured to artificial targets is usually 2mm in 10m or better. The accuracy for all measured points is equal in all three axes.
Depth sensors and inclinometers are used to establish accurately the horizontal plane and hub angles. Hub angles are within ±0.2º.
u Image QC – Initial image QC is done by uploading a test image to check sharpness, lighting and other camera settings. Subsequent images are stored subsea on the camera’s onboard computer
IMCA S 019 19 until the task is complete. A ‘thumbnail’ for every image can be uploaded to the surface or topside computer for visual QC. Final image QC, checking for sharpness, overlap and coverage, takes place on the topside computer. No marking equipment should be moved before this QC process is complete;
u Data QC – Software is used to perform a statistical QC evaluation of the accuracy of measurement of the spool length and the accuracy of hub position and angle measurements. In the three- dimensional photogrammetry model, data is checked against independent readings from inclinometers and depth;
u Limitations – Visibility is a key limiting factor in photogrammetric metrology. Visibility should ideally be at least 3m. At this distance most of the coded targets used on the markers and scale bars should be detected automatically. The visibility should be good enough for the image to cover the width of the spool route marking, and for the targets to be discovered in the images. In low visibility conditions, it may be necessary to position the camera closer to the survey route and take more photographs. Accuracy may not be significantly compromised, but more time and work will be required to process and deliver the results. Processing time is also a potential limitation, owing to the large volumes of data involved.
5.5 INS Metrology
Inertial navigation system (INS) metrology is relatively new to the offshore industry, and the use and availability of such systems has greatly increased in recent years. The principle involved is to use three orthogonal accelerometers measuring linear acceleration in the X, Y and Z plane, combined with three orthogonal gyroscopes measuring angular velocity likewise. Mathematical processing of the output of these instruments, given initial values for position and velocity, makes it possible to track the position and orientation of a device.
INS navigation is broadly similar to dead reckoning, except that instead of using a constant velocity, heading and elapsed time to compute position, the measured real time changes in velocity and attitude are used to compute a real time position. The actual combination of sensor outputs and position computation is very sophisticated and is based on both inertial sensor propagation algorithms and Kalman filtering. Inertial navigation systems are self-contained and do not need an outside signal or external reference to compute a position. However, the biggest drawback of inertial navigation systems is that without external references they are subject to cumulative errors; small errors in the measurement of acceleration and angular velocity are integrated into progressively larger errors in velocity, which are compounded into still greater errors in position. This INS drift increases with the time since an external reference position was last input. In order to maintain accuracy and mitigate these cumulative errors, INS technology offshore is generally used in conjunction with other positioning systems in a hybrid or aided form. Data input from existing positioning systems is used to augment INS data to provide a more robust and accurate overall positioning solution than would be possible with the use of any single system.
More advanced military specification gyros and accelerometers with smaller drifts and biases are becoming available for civilian applications. However, most INS used in offshore metrology applications have a drift rate of the order of one nautical mile per hour. Such systems are referred to as ‘navigation grade’ systems.
INS can be shown to demonstrate the required levels of accuracy for metrology in spools up to 85m long, though of course a more important limitation will be the time taken for data acquisition in order to minimise INS drift. Further advantages that can be expected are: u a reduction in operation times; u impervious to noise generated by surrounding operations such as drilling; u unaffected by acoustic channel management which can constrain operations in busy field developments; u can circumvent problems with line of sight and poor visibility. Obstacles can be ‘flown around’.
INS metrology has great potential for future development and refinement, and is a potential alternative to conventional LBL acoustic metrology. Both stand-alone and combined or hybrid systems (those that combine acoustic range from a seabed reference station with INS navigation) are available.
As INS technology was developed primarily for defence applications, it should be noted that in some countries its use, import and export can be tightly controlled or even restricted.
20 IMCA S 019 5.5.1 System Description and Calibration Considerations
The sensors used to measure inertial acceleration and rotation are included within an inertial measurement unit (IMU). Most IMUs used for metrology today are ‘strap-down’ IMUs. Older systems may have used gimballed IMUs.
Gimballed IMUs can be very reliable, accurate, and relatively low cost. However, they are mechanically complex and are expensive to maintain and calibrate. In ‘strap-down’ IMUs, the accelerometers and gyros are ‘strapped down’ on the vehicle or device being positioned and software is used to keep track of orientation. This method reduces the size, cost, power consumption and complexity of the system.
An IMU consists of:
u three accelerometers that directly measure acceleration in three orthogonal axes;
u three gyros that directly measure rate of rotation in three orthogonal axes.
The outputs that are propagated through the navigation solution are the change in rate of rotation and the change in acceleration. At the system design level many parameters are critical to precise inertial navigation. These include, but are not limited to, the rate at which the IMU is sampled, the method used to translate this data to the required navigation reference frame, temperature compensation, the quality of the gravity model and orthogonality misalignment.
The gyro component within the INS usually defines the term used to describe a given INS. The leading technologies used in subsea INS equipment are fibre-optic gyro (FOG) and ring laser gyro (RLG).
A typical INS consists of:
u an IMU;
u navigation computer to calculate the gravitational acceleration (not directly measured by the accelerometers) and process data from the sensors within the IMU in order to compute a position;
u user interface;
u power supplies.
The unaided navigation solution delivered by an INS is referred to as free inertial navigation . It is subject to drift, which will increase the longer the INS runs without any position correction. A number of different methods can be used to control this drift. In many cases the data from the sensors is combined with or augmented by external data. The most common techniques in use are:
u Doppler velocity log (DVL) – The DVL provides body reference frame velocity data that can be used to constrain the position drift of the INS solution. Very precise alignment between the DVL and the IMU has to be calibrated and maintained;
u Depth – The gravity model of an INS solution usually causes a slightly larger error in the vertical than in the horizontal. A precise relative pressure transducer can constrain this vertical error if integrated correctly. Variations in seawater density during the survey can affect the value measured by such sensors as a variation in density will translate into a perceived variation in depth;
u Acoustic ranges – Acoustic ranges can be used to constrain the position drift of an INS solution. In this case the relative station co-ordinates of transponder locations will need to be accurately determined as well as the depth of the transponders and the speed of sound in water. A variety of INS metrology solutions is available which use additional acoustic ranging. These are sometimes referred to as ‘hybrid’ solutions or ‘sparse LBL’;
u Zero velocity update – If an IMU is held stationary with respect to the earth – i.e. has zero velocity – then the system software, allowing for the earth’s rotation, can remove the drift or error in the accelerometer or gyro sensor data. This method of drift control is very powerful but requires the unit to be stationary for a period of time which will depend on the software configuration and many other variables. Typically this might be for 10-30s or potentially for several minutes.
Before an INS survey can be carried out, calibration or alignment needs to take place. This process consists of:
u Levelling – The accelerometers within the INS are used to orient the system with respect to gravity;
u Coarse alignment – The gyros are used to find the direction of earth’s rotation;
IMCA S 019 21 u Fine alignment – Once coarsely aligned, a very precise estimate of all biases, scale factors and other elements within the Kalman filter are computed.
If sensor aiding is used the reference frame of the sensors used should be aligned as close as possible to that of the IMU. Any residual misalignment can be computed using a Kalman filter.
5.5.2 Equipment List
A typical INS metrology system may consist of:
u the INS sensor package including aiding sensors; this system can be connected to surface through a dedicated ROV data port. The data can be seen in real time and logged on a positioning computer. Power for the INS sensor is generally drawn from the ROV;
u mounting hardware – stabs or docking frames;
u INS online computer;
u INS offline processing computer.
5.5.3 The INS Metrology Computation
The end result of an INS metrology is a listing of positions, depths and attitudes. Each segment of the survey should be logged as an independent data file. Depth will be resolved using either the inertial navigation vertical position solution or by using a pressure sensor in a depth loop. However, if the INS system is depth aided or if the pressure sensor is run simultaneously to the INS, then the loops are inherent in the INS positioning loop procedure.
If the INS is depth aided then a tidal correction method should be used as part of the data processing. All the data should be cleaned for spikes and ‘bad’ entries and then optimally smoothed to compute the final positioning loop results. An average and standard deviation for hub or observation point location is computed.
If the metrology observation point is offset to the hub, the dimensional control offsets along with observed structure attitude must be used to compute the hub datum co-ordinates. See Appendix A.
5.5.4 System Accuracy
It can be difficult to find published values for INS system performance and accuracy that have been benchmarked against a standard or based on a rigorous statistical methodology. However the operator can compare direct measurements using two LBL transducers ranging between themselves. This provides a check for the INS distance measurement and therefore confirms the system accuracy. Once confidence has been established with the system no further checks should be required.
5.6 Subsea Metrology Systems Compared
Every spool design is different, and hence every metrology project is different. There may be one or more metrology techniques which provide an optimal solution. There are a number of influencing factors, including required spool metrology accuracy; water depth; vessel availability; costs and client preference. A table which summarises the main advantages and disadvantages of the main metrology methods described in this document is found at Appendix C.
22 IMCA S 019 6
Subsea Metrology Deliverables
6.1 Computations
The main elements of a spool metrology computation are outlined in the following table:
Horizontal Each hub position is computed from its associated observation point co -ordinates, position of the determined from the INS positioning loop, from the three-dimensional hubs photogrammetric model or from the LBL least squares adjustment. The hub co-ordinates are computed from the OP datum co-ordinates by adding the attitude rotated dimensional control offsets. The co-ordinates should be reported to at least centimetre precision. Depth of the hubs The depth of the hubs should be computed from the depth determined for OP datum. The attitude rotated dimensional control vertical offset is then applied to compute the hub depth. The depth co-ordinate should be reported to centimetre precision. It is good practice to compute a hub-to-hub depth difference, relative to the datum hub. A positive difference is a deeper datum hub, negative, a shallower datum hub. Depth of seabed A seabed profile survey is an important part of any metrology survey. The seabed along intended profile is normally computed in absolute depth; however computation of the relative spool route depth difference to the datum hub is also good practice. Reference should be made to any compensation due to tidal changes at the time of the survey. Attitude of the Measured attitude at the OP is combined with the pitch and roll dimensional control hubs offsets. Both sets of attitudes need to be in the same reference frame. Rotation should be carried out using a co-ordinate rotation matrix. The hub attitude is then reported in the desired heading by a further rotation. Pitch, roll and heading should be reported to decimal degrees. Heading is normally reported in grid. Clarification may be required whether heading data is grid north or true north. Pitch and roll convention should also be clearly specified. Hub-to-hub slant These values are directly computed from the computed hub co-ordinates. Normally and horizontal reported in millimetres. range Spool azimuth The bearing of the spool computed from the hub co-ordinates. Normally reported relative to datum hub. Angle of the spool These parameters are calculated normally only for horizontal spools, they are defined approach as the angle difference between spool azimuth and hub headings. The standard terminology is to call the alpha angle the difference between the datum hub and spool azimuth. The beta approach angle is relative to the opposing hub.
Table 4 – Main elements of the subsea metrology calculation
IMCA S 019 23 6.2 Reporting and Documentation
Subsea metrology reporting requirements may differ depending on which metrology method is used, and should be agreed upon beforehand by the client and the service provider. A typical subsea metrology report might contain the following: u computed spool dimensions: – hub-to-hub horizontal distance – hub-to-hub slant range – hub-to-hub depth difference – connection point attitudes – spool azimuth – spool approach angles if spool is horizontal; u computed hub XYZ co-ordinates and attitude if the methodology resolves them; u details of computation method depending on client requirements; u all recorded survey data in electronic format if applicable; u appropriate drawings/charts. A typical metrology chart should include the spool name, drawing numbers, author, date, client company, service provider, and may consist of four panels: i) a plan view of the subsea structures and the intended spool route, either a schematic or using real co -ordinates, showing the spool azimuth, horizontal true distance and the approach angles (if a horizontal spool) ii) a depth profile of the spool route with the depths of the hubs relative to the seabed. The depth difference is also normally indicated iii) a schematic diagram of the hub attitudes with the attitude convention clearly stated, both graphically and numerically iv) an information section or key containing a summary of spool dimensions and hub attitudes, co-ordinate and attitude conventions, and information such as map projection name and datum, measuring system (i.e. metric, imperial), vertical datum and other geodetic information.
Once the report is delivered to the client, this is only the start of the process of fabricating and installing the spool pieces required. A typical outline of the steps in this process might be: u report issued offshore to client (contractor); u checking of metrology results by contractor survey representative; u metrology results issued to project team, and thence to drawing office; u drawing office produces scaled layouts, from which are produced isometric drawings; u isometric drawings checked by project team and passed onto fabrication contractor; u fabrication contractor makes spool piece (dimensional control, setting out etc.); u on-site checks by contractor quality assurance/quality control engineers; u final gross error check on fabricated spool piece; u full dimensional control survey and as-built drawings (preferable, but not always done).
An example is shown at Appendix B.
24 IMCA S 019 7
References and Further Reading
u Chen C-T and Millero, FJ. Speed of sound in seawater at high pressures , Journal of the Acoustical Society of America, 1977 u Fofonoff, JR and Millard, RC. Algorithms for the computation of the fundamental properties of seawater . UNESCO technical papers in marine science. No. 44, UNESCO, 1983 u Ghilani, CD and PR Wolf. Adjustment Computations. Spatial Data Analysis . John Wiley & Sons, INC, 2006 u Grewal, MS, LR Weill and AP Andrews. Global Positioning Systems, Inertial Navigation and Integration . Second Edition. John Wiley & Sons, 2007 u Leroy, CC and F Parthiot. Depth-pressure relationship in the oceans and seas (J. Acoust. Soc. Am) 103, no. 3 (1998): 1346-1352 u Pike, JM, and FL Beiboer. A comparison between algorithms for the speed of sound in seawater . The Hydrographic Society, Special Publication, 1993 u Wong, GSK, and S Zhu. Speed of sound in seawater as a function of salinity, temperature and pressure (J Acoust. Soc. Am.) 97, no. 3 (1995): 1732-1736 u Further reading: – IMCA S 013 – Deep water acoustic positioning – IMCA D 014 – IMCA international code of practice for offshore diving – IMCA R 004 – Code of practice for the safe and efficient operation of remotely operated vehicles
IMCA S 019 25 26 IMCA S 019 Appendix A Dimensional Control Requirements for Metrology
The dimensional control surveys are normally carried out using conventional land survey techniques based on electronic distance measurement (EDM). A standard resection survey methodology should be used, where a reference co-ordinate system is set up using five or more visible control points and at least two distinct set up locations. A least squares routine is then used to determine all points of interest on the structure, primarily the spatial and attitude relationship between the hub and the metrology observation point. Circle fit methodology should be used to determine best fit centres for hubs, receptacles, buckets, etc.
Before the dimensional control survey is established, a clear definition of the following is required: u the native co-ordinate system of the structure; u metrology observation point or instrument mounting frame native co-ordinate system; u attitude convention.
It is advantageous to define these reference systems from the dimensional control survey and adhere to them throughout metrology operations. This standardisation reduces the risk of operator error.
Overall and individual observation root mean square (RMS) from the best fit solution should always be provided and should not exceed 5mm. The dimensional control survey results should be presented in a format that simplifies the metrology computation. Normally this format is an offset from metrology observation point to the hub in the metrology observation point natural reference frame. Because we observe all measurements at this point, it is then straightforward to rotate the dimensional control offsets to the real world using observed attitude. Computing the hub co-ordinates in real world co-ordinates is then simply an addition of observation point co-ordinates and rotated dimensional control offsets. The same applies for pitch and roll.
Location X (m) Y (m) Z (m) Heading (º) Pitch (º) Roll (º) Metrology receptacle (central) 1.730 -10.676 1.780 -1.86 +0.19 -0.36 Metrology receptacle (NW) 0.000 0.000 0.000 0.00 0.00 +0.00 Structure centre (base) 2.494 -5.407 -3.763 - +0.56 +0.04 12” hub face (W2) -0.470 -6.685 -2.351 268.86 +0.71 +0.38 12” hub face (W4) -0.505 -4.261 -2.320 269.19 +0.62 +0.97
Table 5 – An example dimensional control report with observation point as datum
The dimensional control survey report should always determine heading, pitch, roll and distance from the metrology observation point to all points of interest on the structure, even structure datum. The report should have a clear and traceable computational sequence on the resection least squares results, circle fit computations, attitude computations and any other technical issues. The reference system and attitude convention should always be clearly stated.
A1 Rotation of Dimensional Control Offsets
The complexity of the metrology computation depends on the geometric relationship of the observation point to the hub datum. The dimensional control offsets from OP to hub are normally presented in the structures natural co-ordinate frame, normally with structure level and heading north.
Once the structure is installed on the seabed, the terrestrial reference frame (TRF) is the simplest reference frame we can measure relative to; thus the structure’s attitude measured with an IMU defines the relationship between the structures reference frame and that of the earth.
Dimensional control offsets must be rotated to the TRF by the angles defined from measured pitch, roll and heading. The rotation should be computed using Euler’s rotation theorem or quaternion rotations:
X1 = H m Pm Rm X0
IMCA S 019 27 Where X0 = ( x0, y 0, z 0) are the original dimensional control determined OP to hub offsets in the structure reference frame, Hm is the heading rotation matrix, Pm is the pitch rotation matrix and Rm is the roll rotation matrix, as observed by the attitude measurements and X1 are the new co-ordinates in the TRF. The order of rotation is important. The dimensional control offsets must first be rotated for pitch and roll and then heading. The heading rotation must be last so that the original pitch and roll axes are not changed. Defining a standard co-ordinate convention and making sure that all measurements, offsets, rotations and computations adhere to this convention is critical.
28 IMCA S 019 Appendix B Typical Subsea Metrology Diagram
IMCA S 019 29 30 IMCA S 019 Appendix C Comparison of Subsea Metrology Systems t c y e o r g r t i e
o d l t k
c e r o i r t d f a r i o a t n h r i c a e c r w o e
/
r p y p r r e m e o
e s l o e o o - y r r v b s s s t e r p r o a s s c v
f u p e n e e t e r s e j ) a v c c e t i
n
u
4 n o o u o o h c s s e
i r o 1 r r
c o a o t d t i
p o
0 p p e s
t
n t
l y d n
a u t e m e i a a e r l e r t D t t t r a o i n
e n a a a o a n r c e i l n y t p u r o a p d d A r e
o e a q o e n e s C e e e e e
v c e r v p n r e n n n n n i d M e e o u u h I s O O O O t ( s p b D S O A r e P u u u u u u u u , e t r r s D e o r y u k d t A t s , s e r i n p a r s a e e d c C a a
e
e b v v b n o i i t t d
l a h i t o m e a t t u k e g s i n
s i c p r v p g n e ( e u b s s s i
i w
o j r
p e
r u n n m
d g a r d r r a w a r a u h g n o n o s t
e i n
r r , e i s l o r w o j c r
u l t t t t a s t a
z o o y o
h f l d e o o t u e y e e s c g c m d
o e s s i n o i o p
e
t m n g r n s t a e e m t p m
i v a e e
m
d
e
s
a r j o e a l m b l t r l e n
g b s r a y r o u a n d u e a s o
n f u i e e e n o g e c o o r l
e v i o i n s c , e n b r r r e
r
r i i i t s
o c s e t c e s r
o l z u e g o
l p s a r g t e e a n i m n t e
u l s
w w w p c i o e s n l
t l e a c
t n n e r o s
i
p i s a r i i i ) e e i a d c t t t k s s f e V a e l l D . t d r d n o t t r f l m u r r i n c u u u e o n u n n f c e i s a a T d h O p r r a a l t q a a a u p w o e r i n p p h p t e s T T Q C D R S M O O O p e T T T S F I A u q E u u u u u u u u u u u u u u u u u u s d r e e P e v i x d p d t r r O e d s e l
n
n
o o e f r f p a i d d
e e y )
o e o m e e u b b h c
i m l l s P
e ’ m e t l l i q u u s t t n g
e o v p
o o e o e h h a i i O h g n c n i a
r r r t (
n g d u t
a d
t t
i o i a
o o s t r l t l n q r
i s e n n t t i a
a l
e
n e e f u d
f e b o o i g
t r s s d w y n s o a c c
o s s o a n d
g
e i e s b o l t e e p n y y r s e w e p o
a u n l l c c
n a l s
l l s n e d
k n n c c e i a a o i
e n o
l i d s e a a o o ‘ r v c n n n o
i i y h m e t
t t t s s s s o o r o o g m d g e r
e i i i a a r r g e e e a t a u r n t g s s o v v r r r s p t p n
c m i i i
u u o n i n n
r r s
i n h e s r e u u u y a d e e o j e e e r l a l e c a t s o q q a q s s t c a b v n n
u e a m m t h e e e e b f b i i e e d i u a o o n d M D T S R b m d R L R o O R O s i D u u u u u u u u u u u t a d e y o e r c r t i o g l a t d t
u s e l s s e e s n r u o r v i e n s e a o e c e u e b d g l o c t h e n q
a s a n n p n m
n i i s
e o r i u
e t
r s t r r y L e l
t
s o c o p t u t m d B m o
e t
n n e n n L e o c n e c m d a l
r
a e e r v
a ’ o f
i n d b n u t r r
i t a m s m a d n m t
t s l h
a p y a i r e l p u n e
t n g e i a i i s e o e o a c d n e o l u s f g v r t u
n e
c e r p m
a q s a f o
m r e
e l y e e e e l e i
d y t l y m o y
g
r o p b l d
b d c e e p o i t
e r a i l e e e f v l e o t d t t i x n n t i u p p r r d n s n e i o i t i a a e e q i x a l w u a L e F S D L W G C ‘ C F S E w d m v d A u u u u u u u u u u u u e r i c w i
t t s u u a o t d
c o r a
h e t L v i e B M L D
IMCA S 019 31 s r y l e o r e t h f e n t a t a
r o s r e r i r
e k t a r e i r o p o a h v s t g o o c w
i o s /
s v c t w t e r r r
a c n u o o y e o n l a e s o s
y t h l s r e s
t f e f f f m a l e p e
i k a v
e e e a p c t r c i i i r s i t r t r
i e o o s h h h u o u r g e v r o c c c s i q o t
w p
n o p
e i d p
y y y y c /
s t t t t e a
t t t g r s t r r r C y h n n n o l a o a a o a t l e e e l
e y p Q p p p d i i i f o e r
e c c c k r a i i i n e e e e e e e v r t f f f s v f f f n r
n n n n n n e o o u u u o u o s w s t O S O O S O c m S O O r e u u u u u u u u u P g r n i g o h h
m k n c s s i c o a a b d l r V o e f f a u
d l t O t
d d s c a R
y e n
n
r r g g i – r a r r d e e
n e h
e o i e t t e t o l r e t e g l t i s i
u u l r g w n o n s r u a i a p p a w r w l e o e t t e p
t k s p c s o v w t m m h e s m i c
t s m m n o g u d i o o o s a i e p i d r r e r i o m a l - n r p c c p r t
t a o
c p s u m
e y f u
o s
e s e
l h
r e e r a q b s t y o t
e e e n a r g
p n n
r o e s n a e f t g i i
n n e i h
s n , i d l l s
e i i e i V f t n t s l l r g s a t g g e m i t n f n s f
r i e r t n f f p t
n u n n e O o o o i h u m s d i a e g n
s o o o o p e t
s n R k e r k
r i n u e e
a p t r l i o o m c e S S S m m d p e n n s a o e c e a d p p o n w w o i p y h i n N N N o t a a d c C M I D I M I I e m s T T O O u q u u u u u u u u u u u u E y l t i o l r d d i P P d d g l o r e n o b y n n a i n p i r l a o r r i r O O i a a r s s n
r s n i
e
o o e u p r t t o d t d d
v n o
i h y q
c c e
e m e e u t b b e s c e l
i l l i e ’ t r r d e e l l l c p r u u t n t r n e t i i r r n e e
o
l o f w i i o o h h e a h v t t l i u
e
m a r i
r r o d d g r d t
e u u t s t t t q i m o g o o t n n r m i n d a s e n p p n
i i n n e t t c l a
e
o d u
r r n d
f o o o u S
r s s
m m s s p a d d y i i c o o c c o s s s d f u
o o r
g
x e m s n t N
t t s e e s n I g c c y y r r a t e e i u o n l l
a
c c
n n a l
i
l l e e c n n
c c e – n n e e i y a a o e
o d o p t l
e s a a n g g ‘ r
g r
o o n n n o
p i m n e m m e
t t n n
y o
t
m s s s o i i o o i e u l g e s s s e e r
l e c i i l i t s s e m e e e e v t a r r q l s s i d d o a i o o s s a r r r r v t m o c t m a r l i i i i
u u i e r s n n n n h i n l e e
r
n t t p s s e r t e u u u u u n e e i t e e j r a c c c l h
a e a n a a g c t q q q l q e t u n p p e e b v c o o n n e e m m s t t o n c e e e e e v e e i i r r q d i p u e o a n n e a c o c i d p i l d p d M E C R S S A T M R d R L R R O s i D u u u u u u u u u u u u u u u y e l - s r r n t a o c y o o o
o c e e n
t r e e v
t i m o v v r e s i i i n d o i v t s t t u
i r a s a a e o f s d l l l
e c p g e n t u e e s d
c v r r n e c n m e l n i a l n n
i t i
e s a t a o e e g o o n
s e i i i s c b b t t m o t t
r
e l
e i t a n n a e y c r n n u o r n v
e e g t o r a a m q r u e e t r o c c r n m m s l e r e
e p e m r y y a o s a o
g g f S
e p r o o b m a t n n i m l l s n t
i i t i N o o u e e e e p p m
s I e a e
n n f
t t t r l
n q e e e r d t o o
a u u e o u m p g
i i e d d l l i c o s
s
n t t t
a h
t i i i f s e t e o o y e l t t t a t i t s s b b r d
s s s s s n u o e o i m r u u a y o o o a a w b b a i o b h a a M L P D h F E P N t F S D m v d A u u u u u u u u u u u n y o i e r t r t i a e g i w
m v t a m u n a a
l t r s
a g d l i m t a o o t r t e h i t t e o g s t i e h y n P I s M D
32 IMCA S 019