FEASIBILITY AND DESKTOP STUDIES REGARDING HA 04 FINAL 03.1 Summary and Recommendations

Transcription

1 FEASIBILITY AND DESKTOP STUDIES REGARDING HA 04 FINAL 03.1 Summary and Recommendations P repa red f or: P repa ra t ory C om m i s s i on f or t h e C om preh e n s i v e N u c lea r - Te s t - Ban T r e at y O rg a n i z a t i on ( C T B T O ) P rov i s i on a l T e c h n i c a l S ec ret a ri a t C T B T O, Vi e nna I nte r na ti o na l C e ntr e P. O. B ox , W a g ra m ers t ra s s e 5 A V i e n n a, Au s t r i a P repa red b y : M a l l i n C o ns ul ta nts L t d. 330 T em pe C res c en t N ort h V a n c ou v er B. C. V 7 N 1 E 6 C an ad a 12 A pri l 2012 Summary and Recommendations - Page 1

2 GLOSSARY: AC Air Conditioning ATP BMH CAT CD 1.0 CD 1.1 COTS/NDI CRF CRL CRS CTA CTBT CTBTO DASS db dbv/µpa DDFI DSA DSS ESA FAT Acceptance Test Plan Beach Manhole Commission Acceptance Tests Continuous Data Format and Protocol Formats and Protocols for Continuous Data (CD1.1) Commercial Off-The-Shelf/Non-Development Items Central Recording Facility Certificate Revocation List Cable Route Survey Cable Termination Assembly Comprehensive Nuclear-Test-Ban-Treaty Comprehensive Nuclear Test-Ban Treaty Organization (Commission) Data Acquisition and Storage Segment Decibel Decibels with respect to one volt per micro Pascal Digital Data Formatter and Interface Digital Signature Algorithm Digital Signature Standard European Space Agency Factory Acceptance Test FIT FMECA GCI GPS HDAS HOT hrs. Failures in Time in units 10 9 Failure Modus, Effects and Criticality Analysis Global Communications Infrastructure Global Positioning System Hydroacoustic Data Acquisition System Hughes Olivetta Telecom Summary and Recommendations - Page 2

3 HVAC HVPS Hz IDC IMS ISM M M(Number) MLDT MTBF MTBCF MTTR NDC NDI NMS OSP OTDR PAD PDR PEP PLGR PLIB PMDR PRC PTS QMS RH ROV Heating, Ventilation and Air Conditioning High Voltage Power Supply Hertz (Cycles per second) International Data Centre International Monitoring System International Safety Management Meter(s) Milestone(Number) Mean Logistics Delay Time Mean Time Between Failures Mean Time Between Critical Failures Mean Time To Repair National Data Centre Non Development Item Network Monitoring System Outside Plant Optical Time Domain Reflectometer Packet Assembler Disassembler Preliminary Design Review Preliminary Evaluation Process Prelay Grapnel Run Postlay Inspection and Burial Pre-Manufacturing Design Review Primary Reference Clocks Provisional Technical Secretariat Quality Management System Relative Humidity Remotely Operated Vehicle Summary and Recommendations - Page 3

4 RPL Route Position List RWG Route Working Group SAT SERNAP SIT SLD SLLI SO SOFAR SOH SOR SOW SSI ST System Acceptance Test Chilean Fisheries Service System Integration Test Straight Line diagram System Load and Lay Instructions Station Operator Sound Fixing and Ranging State Of Health Specification of Requirements (Part II of TOR) Statement of Work (Part I of TOR) Standard Station Interface Shore Terminus TOR Terms of Reference TS UPS UTC UWS VSAT Technical Secretariat Uninterruptible Power Supply Universal Time Coordinated Underwater Segment Very Small Aperture Terminal Summary and Recommendations - Page 4

5 FEASIBILITY AND DESKTOP STUDIES REGARDING HA 04 FINAL 03.1 TOPIC 1 Logistics, Environment and Administrative Requirements P repa red f or: P repa ra t ory C om m i s s i on f or t h e C om preh e n s i v e N u c lea r - Te s t - Ban T r e at y O rg a n i z a t i on ( C T B T O ) P rov i s i on a l T e c h n i c a l S ec ret a ri a t C T B T O, Vi e nna I nte r na ti o na l C e ntr e P. O. B ox , W a g ra m ers t ra s s e 5 A V i e n n a, Au s t r i a P repa red b y : M a l l i n C o ns ul ta nts L t d. 330 T em pe C res c en t N ort h V a n c ou v er B. C. V 7 N 1 E 6 C an ad a 12 A pri l 2012

6 Topic 1 - Logistics, Environment and Administrative Requirements Contents Topic 1 - Logistics, Environment and Administrative Requirements Logistics Crozet Île Amsterdam Environment, General Meteorology Environment, Crozet Geologic and Tectonic Setting, Crozet Bathymetry, Crozet Currents and Tidal Streams, Crozet Sea and Swell, Crozet Climate and Weather, Crozet History Forecasting Analysis - Background Analysis - Methodology Fishing, Crozet Anchoring, Crozet Flora and Fauna, Crozet Administrative Requirements, Crozet Île Amsterdam Topic 1 - Page 6

7 6 Conclusions Logistics Environment, General Environment, Crozet Environment, Île Amsterdam Administrative Requirements Topic 1 Appendix A Topic 1 Appendix B Figure 1 Dock in Baie du Marin showing the R/V Marion Dufresne at anchor... 6 Figure 2 Offloading packages in Baie du Marin showing sandy foreshore... 6 Figure 3 The Quay in Île Amsterdam... 7 Figure 4 The Quay in Île Amsterdam in less favourable weather... 7 Figure 5 The Quay in Île Amsterdam from offshore showing rocky foreshore... 8 Figure 6 Île Amsterdam quay and R/V Marion Dufresne standing off... 8 Figure 7 Map of Île de la Possession Figure 8 Crozet monthly wind maxima Figure 9 Consecutive days of less than 25 knot maximum wind Figure 10 Consecutive days of less than 20 knot maximum wind Figure 11 Periods with wind less than 25 knots, Figure 12 Periods with winds less than 25 knots, Figure 13 Periods with winds less than 25 knots, Figure 14 Map of fishing and anchoring exclusion zones as per ARRETE N Figure 15 Detail of navigation chart showing anchorages Topic 1 - Page 7

8 1 Logistics The Crozet Islands and Île Amsterdam belong to the French Antarctic and Austral Territories (TAAF), along with the islands of Kerguelen. Administratively, these territories, together with Adélie Land on the Antarctic continent, come under the authority of the TAAF, an organisation run by the French Ministry of Overseas Departments and Territories. The only inhabited parts of these territories are the scientific bases, which are staffed by 16 to 60 persons depending on the site. The staff is composed, principally, of research personal for geophysical, biological, deep-sea fishing, meteorological and ornithological observatories of research. The Crozet Islands are situated 2,850 km to the south-southwest of the French island of La Réunion, which serves as a back-up logistic base for the three scientific bases on islands belonging to the TAAF. The Crozet base, the Alfred Faure or Port Alfred research station, is the only settlement on the island, is home to between 15 and 60 research personal involved in meteorological, seismic, biological and geological studies. The closest significant port to Crozet is 2,800 km away at Durban in South Africa. Île Amsterdam is situated 2,780 km to the southeast of La Réunion. The Île Amsterdam base, the Martin-de-Viviès research station, first called Camp Heurtin, then La Roche Godon, and the only settlement on the island, is home to about 30 research personal involved in biological, meteorological and geomagnetic studies. Durban in South Africa is located 4,300 km away. The closest significant port is Perth, Australia which is 3,500 km east. There is no airport anywhere in the TAAF region. 1.1 Crozet The Crozet Islands are uninhabited, except for the research station Alfred Faure (Port Alfred) on the East side of Île de la Possession, which has been continuously manned since The other islands are nature reserves only visited during planned scientific campaigns, generally at intervals of several years. The Alfred Faure base which is located at the eastern end of the island on a plateau 143 m (460 ft.) above sea level. There is a 1.6 km road that connects the research station to the coast. The territory is an ecological and ornithological reserve. Any new installation would require an environmental impact study submitted for approval by the TAAF and the Scientific Committee of the I.F.R.T.P. (French Institute for Polar Research and Technology). However it is understood that a repair of the existing Crozet HA04 Station would not present significant permitting issues. A characteristic of all the Austral islands, and a complication for the logisticians, is that there are no natural ports or airstrips. The Alfred Faure base has VSAT communications, and can accommodate up to 10 visitors 1. Crozet is visited four times a year by the 120 m R/V Marion Dufresne, bringing supplies and rotating crews of scientists at the permanent base (Alfred Faure) on Île de la Possession. These visits generally occur in the Austral summer, and are part of oceanographic cruises in the Antarctic Ocean which also include supply runs to the two other island bases of the TAAF. The round trip takes about one month and the transit to Crozet is five days. Crozet is also visited once a month by a patrol vessel, and once a month by long liners of around 55 m in length. Visiting vessels anchor in the designated anchor area where there is some protection from the sea state and the weather. There is a small dock where a lighter can come alongside, and a crane to offload packages. All transfers must be done by lighter or the helicopter on board R/V Marion Dufresne, which can lift 1 ANNEXE A LA LETTRE TAAF/DAIMA-11-n 189 du 15 novembre 2011: Informations nécessaires pour le mouillage d une station hydroacoustique sur les sites de CROZET et AMSTERDAM Topic 1 - Page 8

9 750kg. 2 The water depth at the dock face is unknown, but appears to be approximately 3 m at high water. The foreshore adjacent to the dock and the seabed around the dock are sedimented and unlikely to pierce a hull in the event that a lighter touches bottom. The dock is in a relatively protected site in the Baie du Marin, and is only occasionally subject to significant waves. Speed boats and zodiacs can also be used between anchored vessels and the dock. Figure 1 Dock in Baie du Marin showing the R/V Marion Dufresne at anchor Figure 2 Offloading packages in Baie du Marin showing sandy foreshore 2 Environment, General The Final Report HA04 6 by the independent experts recommended that this study be undertaken on the basis that Île Amsterdam appeared to offer improved weather and more predictable sea states, and hence lower installation risk and maintenance cost. In addition, the report identified weather as the primary cause for the failure of the 2003 installation attempt at Crozet. 7 6 Section 5.7; The Final Report of the Independent Expert Evaluation of the Hydroacoustic Station HA04, Crozet Islands, France; 7 Section 2.2.4; The Final Report of the Independent Expert Evaluation of the Hydroacoustic Station HA04, Crozet Islands, France; Topic 1 - Page 9

10 It is therefore appropriate to attempt a detailed analysis of available weather data to better understand the difference between the two sites, and the weather constraints at Crozet. 2.1 Meteorology The Southern Ocean is the term generally adopted for the circumpolar body of water lying north of the Antarctic continent. The Southern Ocean is divided into two main hydrological zones by the Antarctic Convergence which is defined as the line along which the cold north going Antarctic surface water sinks beneath the warmer sub-antarctic water. It forms a physical boundary between the Antarctic and sub-antarctic zones and is fairly constant in position. The resulting zones are significant because they determine the properties of the air masses above them and therefore directly affect the meteorology of the Southern Hemisphere. The Circumpolar Trough is a belt of low pressure spanning latitudes 60º S and 70º S. It is present as a result of the large number of depressions that have either moved polewards from mid-latitudes or developed in the Antarctic coastal zone. To the north of the Circumpolar Trough in the sub-antarctic zone there is a wide belt of mainly strong winds known as the roaring forties within which the Crozet Archipelago lies. Île Amsterdam lies south of the sub-tropical high pressure belt and north of the sub-antarctic zone in a region of strong westerly winds. 3 Environment, Crozet 3.1 Geologic and Tectonic Setting 8, Crozet The Crozet archipelago is situated on the Del Cano Plateau, which has an average water depth of 2,000 m. This east west oriented plateau extends for about 800 km and is 300 km wide. Its eastern part consists of a raised plateau (Crozet Bank) at a water depth of 1,000 m, on which the five islands that comprise the Crozet Archipelago are built up. The Del Cano ridge was formed at the eastern branch of the Indian mid-ocean ridge, which is now located at more than 1,000 km from the Crozet Archipelago. Associated magnetic anomalies indicate that the Del Cano Plateau formed ca. 50 million years ago. The two larger eastern islands, Île de l'est and Île de la Possession are separated from the smaller western Île aux Cochons, Île des Pingouins and Îlots des Apôtres (a series of 14 small islands and steep rocks; maximum size 1.2 sq. km, 290 m alt.) by the Indivat Basin. All the islands are clearly volcanic in origin. The oldest ages for Île de l'est and Île de la Possession are respectively, 8.8 and 8.1 million years, but there is basalt of undoubtedly older age. Analysis of magnetic anomalies on the sea floor indicates that the Crozet Plateau formed some 50 million years ago. There is little evidence for extensive glaciation. Île de la Possession is made up entirely of volcanic rocks and derived products. Erosion is very active and rocks are covered by very sparse vegetation (moss and lichens). The island was built up in five main volcano-tectonic phases: Phase I: Late Miocene (>8.7 Ma) characterised by the accumulation of submarine formations later exhumed during phase V; Phase II: Late Miocene-Pliocene ( Ma), which saw the construction of an extensive stratovolcano in the west of the island, then the emplacement of a very dense network of intrusive bodies; 8 HYDROACOUSTIC SITE SURVEY REPORT IMS HA4 Topic 1 - Page

11 Phase III: Pleistocene ( Ma), primarily volcano-detritical sedimentation resulting from the dismantling of the previous massif building with intercalated lava flows; Phase IV: Pleistocene ( Ma), with plateau basalts erupted from an Hawaiian type riftzone striking N 135 (Pétrels and Jannels plateaux); Glacial period producing the characteristic land forms ( Ma), and Phase V: Late Pleistocene (0.4 Ma to 10,000 BP), characterised by subsidence tectonics and localised volcanic activity (Stromolian cones). Possession Island is about 18 km long by 15 km wide, with an area of 150 km². The highest point is Mount Mascarin (934 m) in the southern part, but the panoramic summit is Mount Mischief (921 m) which is situated in the south-western part. The overall morphology is controlled by a major topographic feature oriented NNE-SSW broken up by recent volcanic cones (line of craters). This feature extends from Roche Carrée in the south-west to Cap Vertical in the north, thus dividing the island into two sub-areas, the eastern sub-area and the western sub-area. The eastern sub-area is made of plateaux and ridges sloping with shallow angle towards the coast, separated by wide valleys. This terrain represents the remains of a much larger volcanic structure that was probably circular in outline. The missing three-quarters of this massif have collapsed beneath sea level and can now be found to the west and south of the island. The western sub-area has a very rugged relief and is composed of a tilted compartment containing layers dipping at 40 towards the west (known as "Capuches de Moines"). A collapsed zone is situated between the eastern and western sub-areas, containing the Mont des Cratères and the Grande Coulée. Where the valleys do not come down to the coast, there are basalt cliffs, several hundreds of meters high in the south and a few tens of meters high in the north. These cliffs are eroded and cut up into sea-caves, initially formed as lava tubes, into which the waves rush with great force. When the valleys come down to the coast, they form bays with 1-2 km width. The waters in these bays are several tens of metres deep, while the bays themselves are generally also bordered by cliffs. The back of these bays is the only part of the coastline where there are beaches of black sand. The feet of the cliffs bordering these bays also provide a suitable environment for giant seaweed of the type Macrocystis, which makes up a vegetal mat more than 25 m thick coming right up to the water s edge. The beaches are located downstream from these large valleys of glacial origin, in the bays of Petit Caporal, Hébé and Américaine along the Northeast coast, as well as in Lapérouse Bay to the Southwest, the only beach with pebbles. The sand consists of rolled fragments of basalt and derived minerals. The morphology of these beaches can change abruptly when breaking waves are produced by the rare easterly winds. On the beaches of Hébé and Américaine, there are pebbles of floated white pumice brought by westerly marine currents. To the north-east and east, the beaches of Crique du Navire (Baie du Marin), Américaine and Hébé, are located 2 to 3 m above sea level, sometimes forming a micro-cliff (landing at Baie du Marin). They are cut into by the action of the watercourses. These beaches are probably the equivalent of a level located at +4 m described more fully farther west on Cochons Island and representative of the climatic optimum dated at 5,500 BP. As has been confirmed on other Austral islands, the current period is probably one of falling sea level and rejuvenation of relief by river erosion. The discharge of sand at these bays has resulted in some sediment on the seabed, in particular in Baie du Marin, which accounts for the good anchor holding in the bay. The seabed outside the bay can be expected to be a mixture of exposed rock and sediment. Among these bays, Baie du Marin on the east coast is the most sheltered from the prevailing winds. This bay is situated off the mouth of the Rivière du Camp and forms a black sand beach which is 160 m long and 50 m wide. The beach and the stream bank are occupied, as for the other bays, by a colony Topic 1 - Page 10

12 of approximately 100,000 king penguins. The Alfred Faure base overlooks Baie du Marin at an elevation of 140 m. The bottom topography of this bay seems to be regular but slopes rapidly away from the beach until a depth of about 10 m, and then flattens off towards the outlet at 60 m, approximately 1 km further offshore. The bay is 300 m across at the level of the beach, then opening out rapidly to a distance of 1.5 km between Pointe Lieutard and Pointe Max Douguet. Although sheltered from prevailing winds, it is often affected by cross swells which oblige boats anchored in the bay to leave their moorings quickly. When the rare but very strong easterlies are blowing, large breaking waves come onshore and are channelled in the narrow part of the bay towards the beach. There is a record of some seismic activity in the Crozet Islands region. There are occasional earthquakes reported up to magnitude 5 in the region, but their reported epicentres are about 800 km from Île de la Possession and it is not anticipated that seismic activity of this magnitude will be an issue for the marine cable. 3.2 Bathymetry, Crozet The western part of the Crozet Bank is 60 nautical miles wide, with an average water depth of 500 m. The eastern part of this plateau, forming the Île de l Est and Île de la Possession, is separated from the western part by the Indivat Basin with a depth of 2,000 m. The eastern segment of Crozet Bank, where Île de la Possession lies, is much narrower, being only 27 nautical wide miles with depths of less than 500 m. The northern flank of Possession Island becomes steep seaward of the 150 m isobath, falling off to a depth of 1,000 m in less than 10 km; 25 km off the northern coast of the island a submarine plain reaches a depth of 2,300 m. Figure 7 Map of Île de la Possession By contrast, the southern flank extends onto a wide plateau that slopes regularly to a depth of about 500 m over a distance of 40 km, continuing westward to the Duclesmeur plateau which joins up with the western segment of the Crozet Bank on the other side of the Indivat Basin. Possession Island is separated from Île de l Est by the Canal des Orques, at a depth of m, which connects up with two relatively narrow canyons situated to the north and south. The submerged part of the Île de l Est massif is considerably narrower since it does not extend onto any submarine plateau. Topic 1 - Page 11

13 The 200 m isobath represents the boundary of a 5-7 km-wide terrace to the north of Île de la Possession. Beyond this limit, the slope appears to steepen abruptly down to the 1,000 m isobath. 3.3 Currents and Tidal Streams, Crozet In general 9, south of 40º S, the currents mainly set towards the east and are collectively known as the Southern Ocean Current. The Southern Ocean Current tends to fluctuate in direction between north east and south east with low to moderate constancy, with an average rate of ½ to 1 knot. There is insufficient data to determine whether there are seasonal variations of the current. After prolonged periods of strong winds from a constant direction a wind drift current may be generated where the rate varies according to the speed of the wind and its duration. These wind drift currents may strengthen, weaken or reverse the surface current and cause major irregularities in the set of the current across the region. In the region lying west and north west of Iles Crozet, approximately between parallels 44 S and 46 S and meridians 40 E and 56 E, the general flow is not in accordance with the usual direction of the Southern Ocean Current. While north west, east or south east sets may be experienced the predominant sets are between north west and north east throughout the year. The normal east set is found between 120 and 150 nautical miles farther north. Local effects may also be experienced closer to shore and within the various channels between the islands. For instance during the 2003 installation, during the cable landing currents of over 1.5 knots were experienced in Baie du Marin. It is understood that from time to time a strong current flows between the islands. 3.4 Sea and Swell, Crozet Sea waves 10 are generated locally by the wind and can be variable in direction. Wave conditions in the Southern Ocean throughout the year are similar to those experienced in the Northern Part of the North Atlantic in winter, with the worst conditions likely to occur between latitudes 45º S and 60º S. Within this region (Crozet lying close to the northern limit of this general belt) rough seas are common in all seasons becoming very high or even phenomenal during the passage of the numerous deep east moving depressions. Maximum wave heights are thought to be around 25 m over this mainly westerly wind region but can attain greater heights in some areas (e.g. 35 m near Isles Kerguelen) in winter. The Southern Ocean is a stormy region and therefore there is a very real possibility of abnormally high waves developing. Waves generally become steeper and higher with an opposing current, and there is some evidence that large waves become even higher on the approaches to submarine banks (e.g. the Aghulas Bank). This increase in wave height may be due to the both opposing currents and the effect on the waves as they move toward shallower water. Swell data for summer months is limited and almost non-existent for large areas in winter and therefore should be used with extreme caution. The British Admiralty note that the percentage frequency of swell heights of 3.5 m and above is considered to be higher than indicated in their Swell Distribution Charts and in the region between latitudes 45º S and 60º S may be present for around 35% of the time in Summer and about 65% of the time in winter and are, in the majority, from a south westerly or north westerly direction. Wave data has been provided by IFREMER. The data file received provides a visualization of significant wave heights throughout the world which occurred during the month of January This visualization serves to confirm that storms with wave heights of m do disturb the ocean around Crozet fairly regularly, even during Austral summer. These events are followed by periods of 9 Admiralty Sailing Directions, NP9 and NP39 10 Admiralty Sailing Directions, NP9 and NP39 Topic 1 - Page 12

14 2-4 m wave heights lasting 12 to 96 hours. This data tends to confirm results of the analysis of the Meteo (France) wind data. 3.5 Climate and Weather, Crozet Lying as they do in the path of the "Roaring Forties" (between 45 95' and 46 50' S, 50 33' and 52 58' E), the islands are invariably windy, and the frequent depressions arriving from the west bring cold, wet and cloudy weather. On average it rains 300 days per year, and winds exceed 100 km/hr more than 100 days of the year. The climate is typically Austral with a mean annual temperature of +5 C. The temperature rarely exceeds 64 F (18 C) in summer. Frost or snow is rarely recorded (89 days per year, on average). Most winds are in the north-westerly quadrant, the prevailing direction being westerly and the strongest winds coming from the north. The following is a quote from the Final Report 11 : In generally, it can be said that this phase of the installation was thoroughly disrupted by the weather. The forecasts and weather maps sent to us every 12 hours were very precise and we relied totally on these when initiating operations that could last anywhere from 7 to 20 hours, depending upon the sequence that was to be carried out. Unfortunately, periods of high winds (>35 knots) lasting approximately 4 hours occurred virtually every 24 hours. Following such winds, the sea could be moderate to rough, making cable-laying impossible. It was not feasible to sequence the various operations as planned and various procedures had to be interrupted after the portion of the array already in the sea was secured. No operational sequence could be performed as originally planned without enormous loss of time, attachment of buoys, and the subsequent pick-up of mooring lines. The only period of calm seas occurred between 10:00 hours on 24 March and 20:00 hours on 27 March, i.e. a period of 3 days and 10 hours over the entire period between 15 March and 5 April, these being the dates of commencement and termination of the operation in the South zone, i.e. 20 days. It is important to quantify the duration and frequency of unworkable weather. The duration and frequency of unworkable weather will impact the technology solution, the cost estimate and the risk analysis as they apply to Crozet History From 2000 installation presentation 12 : Date Range: 20 Feb March 2000; Weather forecasting service : 2 forecasts per day; Gales of 45/50 knots (fairly common during the mission) created 11 to 12 metres high waves when combined with low-pressure systems arriving from south; R/V Marion Dufresne : 25 days on site, 3 days declared weather stand-by (12% weather); N/O La Curieuse: 24 days of on site, 3 days declared weather standby. From 2003 installation presentation 13 : Date Range:15 Mar Apr 2003; Reliable weather forecasts and weather maps received every 12 hours; 36 days on site; 11 HA04 Hydroacoustic Station CTBTO N 99/30/6010 Installation Files October 2003 Final Report (2) Version 1.0 Index A 12 Installation of HA04 station , Independent Expert Review HA04 station, Vienna 10/14 May Installation of HA04 station , Independent Expert Review HA04 station, Vienna 10/14 May 2010 Topic 1 - Page 13

15 Weather stand-by: 14 days (wind >30 knots or swell >11 m) (39% weather); CS René Descarte Forecasting During the 2003 installation, a custom weather service was procured for the vessel with 12 hour updates. Weather forecasting was found to be reliable, and was relied on throughout the work. Topic 1 Recommendation 1 Commence discussions with the selected weather forecast agency well in advance of the work, so that the forecasters have the opportunity to test their models against actual weather for several seasons prior to the vessel arriving on site. It is also noted that a Box and Whiskers type of forecast, where the error bars and likelihood are provided with the forecast, could assist the site crew Analysis - Background The intent of the analysis of weather data at Crozet is threefold: To determine the optimal months for the work; To estimate the number of hours of consecutive work time that can reasonably be expected by the Contractor; and To estimate the percentage of downtime that should be allowed for in the cost estimating. It is acknowledged that the outcome of this (and any) weather analysis will always be a greatly simplified solution to a very complex problem. For instance, the ability of a vessel to maintain station in high winds and seas depends on wind speed, but also swell, duration of high winds, and the relative directions of wind, sea and swell and current. A vessel will aim to stem the strongest influencing force, normally the wind in this circumstance, but often the resultant force, with due regard for the sea and swell state, direction and period will affect the ability to maintain position. In addition the master must consider the safety of crew working on deck. The vagaries of the operation also have to be taken into account, in particular when working with cable outboard which may potentially lead under the stern and/or into propulsion machinery. A vessel may not always be unable to do any work during bad weather. During the 2003 installation, work was rescheduled so that tasks with less weather dependency or in areas offering some protection could be done during periods of bad weather. However, despite all of these limitations that apply to a weather analysis, it is necessary to undertake analysis of the data that are available, in this case wind records, in order to provide justifiable approximations of likely consecutive work hours and percentage downtime Analysis - Methodology The analysis of weather at Crozet was undertaken using weather data from the Meteo web site, and the daily reports from the France Telecom (FT) Final Report on the 2003 installation program. The first part of the analysis was decadal, to determine the months during the Austral summer when the weather would be least hostile. This analysis is in Section The second step in the analysis was to try to determine a suitable relationship between land wind speed as reported by Meteo and Sea State scale sea state that could be anticipated for a vessel working at sea. The method used to develop this relationship was to compare the hourly weather reports from the FT daily reports with the Meteo weather data. Unfortunately, the weather station on Crozet did not record the wind speed for much of the time that the FT ship was present during the 2003 installation. However, the data which were recorded were used to make a reasoned approximation of times when a vessel could be expected to stop work based on Meteo wind speeds. The second stage analysis is in Section Topic 1 - Page 14

16 The third stage of the analysis compares daily maximum data for one Austral summer ( ) and compares it to the selected cut-off speed. The intent of this analysis is to determine whether the cutoff is reasonable, and to determine the sensitivity of the analysis to cut-off speed. This third stage analysis is in Section The fourth stage of the analysis was to review hourly data for December-February for several years. The intent of this analysis was to determine the number and length of weather windows that could reasonably be anticipated during any one month period on site Austral Summer The first step in the analysis was to determine the most favourable period in the Austral summer for installation. The Austral summer period at Crozet was identified in the Survey Report 14 as running from December to February, when the risk of weather stand-by was said to be minimal but not entirely negligible. The survey report also observed that certain weeks in autumn or spring can be favourable, with a little luck. An analysis of weather records for Crozet has been undertaken in order to confirm the duration of the best weather period. The data for this analysis was purchased from the Meteo France web site. The monthly data sets for wind velocities and direction were downloaded for the decade 2002 to Two data sets were downloaded, for maximum instantaneous wind speed during each month (FXIAB) and for the highest daily average wind speed for each month (FXYAB). The intent of reviewing this data was to define the optimal weather months for installation. Figure 8 summarises the data and the conclusions of this analysis: that the optimal months for installation are December to February, and that the chance of severe storms with wind speeds over 70 knots is significantly higher in March to November than in December to February. However, the Climatic Table for Alfred-Faure Station, Isles Crozet 15 compiled from 13 years of observations between 1980 and 1995 also indicates that the Austral summer winds have a lower average speed, knots, during January and March (18-21 knots in February) than at other times of the year and an average occurrence of gales on only 7 days of the month in each of the months between December and March, inclusive; the Austral winter average peaking at 14 days in July. For example the average wind speed shows January and March as slightly better than February and December. The prevailing winds are south-westerly to north with north westerly winds slightly dominating over south-westerly although winds from all directions were observed. 14 HA04 Hydroacoustic Survey Report 14 august 1998 CEA/DAS 15 British Admiralty Sailing Directions NP Seventh Edition 2009 Topic 1 - Page 15

17 Figure 8 Crozet monthly wind maxima Topic 1 Conclusion 1 The most favourable weather window for installation activities is approximately December 01 to March 01. However the month of March may also offer weather windows, and if the work extends into March it is unlikely that the weather risk will increase significantly Comparison of 2003 March-April data The second step in the analysis was to determine a reasoned relationship between land wind speed as reported by Meteo and the time when vessels have to cease work due to weather. The ability of a vessel to operate and undertake cable work is determined by the judgement of the vessel s Master, supported by his/her experience and knowledge of the vessel. A commonly used measure of workability of the sea state is the Sea State scale. The Sea State scale takes into account wind speed and sea state. The Meteo data provides readings of wind speed as the wind crosses the island. It is always a challenge to compare land wind speed to Sea State scale sea conditions and to come up with a meaningful prediction of workable time. However in this case there is some vessel data for 2003 that can be compared to the Meteo data. Unfortunately, while there are vessel data for 868 hours, the corresponding Meteo data are very sparse, and there are only one hundred and seventy four hourly readings for the same period. The data downloaded was wind speed, averaged over ten minutes, and reported as the maximum 10 minute average in each hour. This data eliminates short gusts, but is conservative in that ten minutes of wind will not excite waves unless the high winds persist. Of these 174 readings, 59 are during the time that the vessel was delayed by weather, and the remaining 115 are during times that the vessel was working. A review of the wind speeds from Meteo during those times is detailed in Table 1. If an arbitrary limit of 25 knots wind speed is used for the work cut-off, during 14 one hour periods the vessel did not work although the wind speed was below 25 knots, and in 20 one hour periods the vessel worked even though the wind speed was over 25 knots. A review of the 14 hours when the vessel was on standby and the wind speed dropped to below 25 knots were hours during a major storm event; i.e. the adjacent times were stormy, and the sea had not had time to settle. Topic 1 - Page 16

18 A review of the 20 hours when the vessel was working in over 25 knots of wind shows that 11 of the 20 hours were associated with the weather event that resulted in the failure of the N1 cable; the remainder occurred when the vessel was in transit, or when the vessel recorded limited swells. These numbers suggest that a wind speed cut-off of 25knots is reasonably representative of the experience on site. If the wind speed cut-off is increased to 30 knots, the number of occasions when the work was suspended when the weather was less than the wind speed cut-off increases to 26. These periods are still generally during a major storm event. The number of periods when the vessel worked in winds higher than the wind speed cut-off drops to 12. These numbers suggest that a wind speed cut-off of 30 knots is as representative of the experience on site as a wind speed cut-off of 25 knots. cut-off 20kn 25kn 30kn Hours when work stops with wind speeds less than cut-off Hours when work continues with wind speeds greater than cut-off Table 1 Cut-off speeds If the wind speed cut-off is reduced to 20 knots, the number of occasions when the work was suspended when the weather was less than the wind speed cut-off is 4. These periods are still generally during a major storm event. The number of periods when the vessel worked in winds higher than the wind speed cut-off increases to 41. These numbers suggest that a wind speed cut-off of 20knots is not representative of the experience on site. Topic 1 Conclusion 2 Based on this analysis, it seems reasonable to propose a cut-off for vessel work of 25 to 30 knots wind measured on shore, provided that the average wind in a 9 hour period is more than 25 knots to eliminate short duration breaks during a storm event. In order to err on the side of conservative, a wind speed of 25 knots (represented by Beaufort Scale 6 16 ) is used as a suitable indicator of unworkable weather, provided that the average wind in a 9 hour period is more than 25 knots Maximum Wind Speed 10 Minute Period in 1 Day The third stage was to analyse the maximum average wind speed for any 10 minute period in a day 17, reported by the day for the selected installation period. The data for 15 Nov 2010 to 15 March 2011 was downloaded, and an analysis carried out using a wind speed cut-off of 25 knots. This cut-off showed the following continuous work periods of over 24 hours during these 17 weeks: 6% of the time there were 4 full consecutive days of less than 25 knot winds; 10% of the time there were between 3 and 4 full consecutive days of less than 25 knot winds; 6% of the time there were between 2 and 3 full consecutive days of less than 25 knot winds; and 10% of the time there were between 1 and 2 full consecutive day of less than 25 knot winds. To summarise, in 17 weeks of the Austral summer , when considering days from 00:00 to 24:00: 40 days had no ten minute period where the wind averaged over 25 knots; 16 See Appendix A 17 Vitesse vent quotidien maxi moyenne sur 10 min Topic 1 - Page 17

19 for 33% of the days average wind in any 10 minute period was under 25 knots. Figure 9 Consecutive days of less than 25 knot maximum wind To determine sensitivity to the wind speed cut-off, a further analysis was done with the cut-off reduced to 20 knots. This analysis is summarised in Figure 10. This 20 knot cut-off showed the following continuous work periods of over 24 hours: 6% of the time there were 2 full consecutive days of less than 20 knot winds, and 6% of the time there was 1 full day of less than 20 knot winds (but less than 2 full days). To summarise, in 17 weeks of the Austral summer , when considering days from 00:00 to 24:00: 16 days had no ten minute period where the wind speed averaged over 20 knots; for 13% of the days average wind in any 10 minute period was under 20 knots. This sensitivity analysis shows that the analysis is very sensitive to wind speed cut-off, with the number of days with wind speeds below the cut-off halving for a 20% reduction in wind speed. However this analysis is coarse, since it is based on fixed one day periods, and makes no allowance for weather periods that extend beyond one day but less than two days, or for brief high winds that affect a day s total but would not affect sea state. Topic 1 - Page 18

20 Figure 10 Consecutive days of less than 20 knot maximum wind Maximum Wind Speed 10 Minute Period in 1 Hour The fourth stage of the analysis reviewed the maximum wind speed for any 10 minute period in an hour, reported hourly. These data were downloaded for November 1 to April 30 for , and was selected to match the stage 2 data; was selected because there appeared to be storms in Jan-Feb for that winter; and was chosen to validate assumptions by correlation with reports from the 2003 installation as per Section Figure 11 Periods with wind less than 25 knots, Topic 1 - Page 19