Deepwater horse power

As an energy resource, Thunder Horse holds the potential to produce around one billion barrels of oil - BP holds a 75 per cent stake as operator of the Thunder Horse licence with co-owner ExxonMobil holding the remainder. The field, located 240 kilometres southeast of New Orleans in 1900m of water, came onstream in June 2008, at first through a single subsea well, quickly followed by others to rapidly ramp up production. Today eight subsea producing wells are in operation, with more planned.

'At every node of the Thunder Horse development there has been a unique challenge,' Replogle explains. He proceeds to outline just a handful of the many technology 'stretches' that the project worked on with vendors and specialists in an unprecedented collaborative programme of equipment development, testing and qualification to deliver a new generation of engineering solutions.

'The Thunder Horse reservoir conditions are among the toughest encountered in the industry,' he notes, 'producing fluids at pressures over 1200 bar and temperatures up to 135*C, which can also be corrosive due to the presence of carbon dioxide and hydrogen sulphide. In 1999, there was no subsea equipment capable of handling such fluids, hence high pressure subsea trees - the arrangement of valves that control the wells at the seabed - and all other related subsea equipment and control systems had to be specifically developed for the project, equipment that also had to be designed to operate in 1900m of water.

'The depth of the reservoirs means that very long, high pressure wells have to be drilled, some of them reaching almost 9000m into the seabed. Existing drilling techniques and well completion equipment of the day had to be enhanced to new levels to deal with these demands, resulting in the project developing many innovative components in high strength materials for completing the Thunder Horse wells.

The list of major challenges continues unabated. For example, conveying large volumes of high pressure fluids from the subsea wells to the surface required a step change in the design and size of steel catenary riser pipes employed for this. Furthermore, the riser metallurgy has to withstand long term fatigue caused by continuous movement in the ocean currents and the motions of the PDQ, additionally requiring a new type of 'flexjoint' to be developed to support the risers as they connect to the moving platform. Similarly, the size of flexible risers used to carry water to the wells for water injection into the reservoir also required a jump in technology.

And even the geology surrounding the Thunder Horse reservoirs posed unique problems. The field's two reservoirs lie below a thick layer of salt in the subsurface which creates difficulties in obtaining and interpreting good seismic images of the hydrocarbon-bearing formations below (see 'Subsurface challenges' panel at end of article). The leading-edge seismic imaging methods that were necessary to overcome the subsalt problem, techniques developed by BP, have subsequently placed the company at the forefront of the industry in this vital area of expertise (Frontiers, December 2005).

BP has since invested heavily in marine assurance technology and has installed two independent real-time monitoring systems on the PDQ, and on other BP platforms.

'Instruments on the platform continuously measure a range of operational parameters, for example vessel motions, ocean currents, wind speed, mooring line tensions, and the air gap below the PDQ's deck,' says Neil Cramond, marine authority for BP's operations in the Gulf of Mexico. 'The data are relayed to shore either by fibre optic cable to onshore support centres, or via satellite link - the latter system is powered by solar cells and batteries and also transmits photographic and video images of the PDQ.

'The net benefit is an improved assessment of operational integrity on a day-to-day basis, and rapid confirmation of platform performance during hurricane evacuations - BP can now see, in real time, what is happening on and around its unmanned facilities during a hurricane, a capability unique to BP's platforms in the Gulf of Mexico.'

'We made incredible headway surprisingly fast,' notes Greg Rohloff, manager for offshore operations. 'This was impressive when you consider the logistics of having around 700 people working offshore, moving daily between the PDQ and three flotels, with a large fleet of support vessels and with hurricanes coming into the Gulf throughout the season.'

The delay for recovery was used very effectively by the Thunder Horse operations team, led by Sammy McDaniel.

'Our task was to create a world class operational readiness plan,' explains McDaniel, 'covering both the remedial work, hookup and commissioning, and integrated startup, so that when we reached startup we would be fully prepared in all respects. There were 900 actions in the plan, all of which were carefully vetted by experts throughout BP. A major aspect of the plan lay in training the 140-strong PDQ operations crews, in which BP has invested significantly. We had assembled an experienced team brought in from BP's worldwide operations - though none of them had been on a platform as large as the Thunder Horse PDQ before.'

Given the scale and complexity of the PDQ and the large number of interdependent systems to be managed and maintained - incorporating many state-of-the-art equipment items and technologies - highly skilled personnel are required to operate the facility. A purpose-designed computerised training simulator played an important role in testing competence and preparing the crews, enabling the team to 'virtually' rehearse startup and regular operations in BP's Houston office before moving offshore to the PDQ.

All of the field's wells will be completed as subsea wells, some drilled from the PDQ, others drilled by mobile drilling units at distances up to 10 kilometres from the PDQ. The HP/HT wells are sited at seven main 'drill centre' locations around the field, each with a manifold to which the subsea wells are connected by steel flowlines on the seabed. Although wellfluids from the reservoirs are hot, near-freezing sea temperatures on the seabed could cause ice-like hydrates to form in the lines under some conditions, requiring them to be insulated and also to have a chemical inhibitor injected into the hydrocarbons - the project developed a modified low dosage hydrate inhibitor for this purpose.

Five steel catenary risers (SCRs), the most robust of their kind in the industry, take hydrocarbons from the seabed to the PDQ while two more SCRs route the processed oil and gas from the PDQ to the main export pipelines. Risers on Thunder Horse have to be longer and stronger than any before, up to 600mm in diameter with walls thick enough - up to 40mm in some cases - to resist both the internal and external pressure. Three flexible risers will carry treated seawater and produced water from the PDQ to subsea water injection wells - together, the length of risers and flowlines totals around 75 kilometres. To distribute chemicals to the wells and supply hydraulic and electrical control signals, 18 multi-core flexible umbilicals will interconnect the subsea components, totalling over 60 kilometres in length.

The subsequent scope of equipment development, testing and qualification programmes carried out by the project and its key subsea systems suppliers amounts to a major project in its own right with investment running into many millions of dollars.

'To design and manufacture the high pressure subsea trees alone there were 57 technology stretches to be met,' adds Bednar. 'If you consider the technology evolutions for all of the subsea architecture - for example valves, gaskets, seals, couplers, connectors, stress joints, coatings, insulations, control systems, risers and corrosion-resistant materials, to name but some - Thunder Horse can rightly claim to have been the main driving force which has given the industry the know-how to tackle new fields in even deeper water that have higher pressure and temperature reservoirs, some of which are now moving towards development.'

Indeed, the project is responsible for 'creating' over 70 new subsea products, all of which were subjected to a series of rigorous testing stages - including prototype, qualification, endurance, factory acceptance, systems integration and wet testing - before they were included in the subsea systems, combined with a high degree of third party inspection.

One such achievement focuses on the flexible water injection risers, as Karen Veerkamp, subsea engineering manager, explains.

The three smooth-bore flexible risers, designed and built by the project's subsea engineering contractor Technip, are rated to withstand pressures up to 690 bar and have internal diameters of 200mm - concentric external layers of composite materials and steel push the outside diameter to 350mm. The risers, which are supported from the PDQ's hull, are up to 2450m long and weigh up to 635 tonnes.

At present, no water injection is taking place in the field as the aquifer underlying the Thunder Horse reservoirs is naturally maintaining pressure in them, but in time water injection will be needed to keep pressure up.

In mid-2006, as the PDQ was approaching its revised startup date, four of seven subsea drilling centre manifolds with associated flowlines and control systems were in place on the seabed, including the two largest manifolds immediately below the PDQ on Thunder Horse South. These large structures, each measuring 20m long by 7m wide and weighing over 300 tonnes, had been installed by the project's main installation contractor Heerema in 2004 before the platform had arrived on station. Four subsea producing wells were connected to the manifolds - each of which is capable of supporting up to 10 wells - in readiness for field startup.

The unwelcome discovery led to an extensive metallurgical investigation at the most detailed level, and a complex offshore restoration operation.

'At first it was not clear why the failures occurred,' says Veerkamp. 'The materials, welding and post heat treatment methods used in fabrication were those regularly used in subsea applications by the industry, and all fabrication and test procedures had been rigorously adhered to - furthermore, the manifolds had successfully passed the hydrotest when first installed two years earlier.'

While a team of multi-disciplined BP technical specialists brought into the project pursued the investigation, the company took the major decision to remove all subsea hardware that contained the same types of material and welding - the two manifolds alone together contained 582 welds.

'The impact that this would have on field startup was evident, but no risks could be taken,' states Replogle.

'It was very important to understand what caused the failure in order to avoid future risk in refurbishing the equipment,' recalls Jim Burk, BP advisor in materials and corrosion. 'The efforts of many BP specialists in materials, welding, corrosion and subsea engineering, working closely with their counterparts in ExxonMobil, led to an understanding of what occurred.'

It was found that minute cracks had formed in the 75mm thick thermal insulation on the manifold pipework, leading to the disbonding of the multi-layer anti-corrosion coating on the pipe surfaces. By exposing the bare steel pipe to the subsea environment, the manifold's cathodic protection system - the standard method to prevent corrosion around offshore structures - started producing hydrogen at the pipe steel surface and charging hydrogen into the manifold's forged fittings. As hydrogen concentration built up over time, this led to hydrogen embrittlement in the weld interface between pipe and fitting, causing the interface to crack and fail under pressure.

For refurbishing the subsea equipment, a new welding procedure employing different materials and weld configurations was developed by BP, and subjected to intense review and qualification. State-of-the-art inspection techniques, such as phased array ultrasonic technology, were employed, while every detail of the welding operation was strictly controlled - including even the sequence used by welders to create weld beads. BP has since used the experience gained to develop a new set of best practices and standards for application in future HP/HT subsea projects.

While the subsea equipment was being replaced, the operations team created a 'preservation programme' to keep the topsides equipment in operation.

'The topsides have a number of closed loop flowpaths in the design which allow us to circulate fluids around when warming up prior to startup,' says Wissam Al Monthiry, deputy operations manager. 'These loops allowed us to circulate diesel oil, water and chemicals around the production systems, and we also bought natural gas from the Mardi Gras pipeline system to run the power generation turbines.'

As the operation progressed, the team was able to run the entire platform as if in production. The move proved its worth, not only in giving operators hands-on experience with the plant and systems, but also in finding all the teething troubles that typically arise in the first year of running a new facility.

'We encountered and solved many problems before they cost us downtime later,' Al Monthiry points out. 'It also kept people motivated and it proved really valuable when we came to the actual startup.'

'The combination of deep water, salt overhang layers and very deep reservoirs requiring long wells is just one part of the challenge,' says Charlie Holt, wells delivery manager. 'Add to that the high reservoir pressure, the high flowrates - some of the wells can individually flow at 50,000bpd or more - and the fact that hydrogen sulphide can be present in the wellfluids once water injection begins, and you have a drilling and completions challenge that no-one had faced before.'

Hydrocarbons lie in three zones in the Thunder Horse reservoirs, lying between around 5500m and 8000m vertically below the seabed in the Miocene sandstone. When drilling any well, a key parameter is the ratio of pore pressure to fracture gradient. The pore pressure from the hydrocarbons in the rock determines the pressure that must be exerted by the drilling mud circulated through the wellbore to prevent the hydrocarbons flowing into the well as it is being drilled, while the fracture gradient indicates the equivalent mud weight that - if too high - will fracture the rocks, an event to be avoided. Drilling a successful well requires staying in the window between these two controlling pressures, a window which varies with drilling depth.

'That window is very narrow for Thunder Horse wells,' adds Holt. 'The fracture gradient in the relatively young Miocene sedimentary rocks is low, while the pore pressure is high. Consequently we have to control the weight of the drilling mud very precisely to within ounces per gallon, far tighter than on other wells in the world. Eventually the window becomes too narrow even for this measure, requiring us to put another casing string into the wellbore to prevent reservoir fluids forcing their way in.'

Throughout the life of a well, intervention is necessary, from the first perforation of the well to prepare it for production, followed by routine maintenance. The industry's standard method for doing this in long, deviated wells is by coiled tubing, conveying equipment and tools to the end of the well on a reel of continuous small-bore flexible tubing. Coiled tubing can reach the bottom of Thunder Horse wells in the two upper hydrocarbon-bearing zones, but in the lower zone the depth combined with step-out distances of the order of 2500m limit coiled tubing activities.

'To address this, we have developed a solution known as hydraulic workover (HWO),' says Ken Armagost, team leader for reservoir access. 'Rather than coiled tubing, HWO uses jointed pipe passing through a purpose-designed hydraulic jacking unit which will enable us to achieve enhanced and cost effective interventions in the deepest wells at step-out distances over 4000m - this means we can drill even longer wells and these will give us access to more hydrocarbons.'

The HWO unit, which will also improve operational safety during interventions, is built and is being tested now by BP in Louisiana, and could be in operation by the end of this year.

In the subsurface, the wells are completed using a variety of methods for preventing sand entering the wells along with the wellfluids, including cased and perforated methods, high rate water packs and frac packs, all designed to boost production (Frontiers, August 2008).

Given the high flowrates in Thunder Horse wells, large-bore tubing is needed at the core of the well - up to 152mm internal diameter - while the unusually high pressure dictated a new design for subsurface safety valves, located in the wells some 1125m below the seabed. Corrosion resistant alloys, such as nickel-chromium based alloys and 25 per cent-chrome super duplex stainless steel, are required for all critical components in the completion strings, which, due to their length, result in some 225 tonnes of sophisticated equipment being deployed into a well to complete it.

'Many of the components in the Thunder Horse well completions are new for the industry - developing these was necessary to match the demanding operating conditions,' points out Jack Shada, senior completions engineer. 'The project set up an extensive testing and quality assurance programme - of the 32 major components in a Thunder Horse completions string, 18 of these are classed as "serial number ones" - that is to say, they are the first of their kind ever made. Including the methods developed for installing the completions from the sea surface, and operating them, over 100 serial number ones are incorporated into the wells.' (Frontiers, August 2004.)

The project initiated the development and qualification of an alternative surface treatment to be used in the manufacture of new tubing components to reinstate the well - which is now in operation - and for all further Thunder Horse well completions.

Looking a short time ahead, several wells in the field are set to become 'intelligent wells', that is to say they will be equipped with completions that allow downhole control of different reservoir zones.

'At the time of planning Thunder Horse wells, the concept of hydraulically controlling downhole completions equipment in different reservoir zones in the same well was a fairly new idea,' explains Holt. 'We have been working on this technology with our main completions supplier Baker Hughes and it has now moved ahead to the point where we can employ it in the field - the wells were originally designed to accommodate the technology at a later stage. Intelligent water injection wells in Thunder Horse South will enable us to inject into separate reservoir zones from a single well, while in the more remote wells in Thunder Horse North we will be able to produce from multiple reservoir zones in one well without commingling the wellfluids. This will be a significant advance for such deep, high pressure wells.'

'Our target for operational uptime in the first year was 89 per cent,' says Tommy Hassold, one of Thunder Horse's offshore installation managers. 'Everything has run very smoothly such that we have reached over 93 per cent.'

Eight wells are now in production, four in the south and four in the north. More will follow in 2010 and beyond.

Throughout the project, the safety of the wider Thunder Horse workforce has been given top priority, right through the complex offshore operations carried out with hundreds of people on the move and many different vessels working around the field, through the rapid ramping up of production, and now during drilling and production as the project moves forward. Since the PDQ went offshore in 2005 almost 14 million manhours have been expended on the project, with only two cases requiring days away from work.

'This is an outstanding safety record and testament to BP's safety training and procedures,' notes Thunder Horse vice president Dan Replogle.

Looking back over the past few years since he joined the project, Replogle acknowledges the team has overcome every challenge presented along the way.

'It needs to be re-emphasised that the Thunder Horse project was developing prototypes on many fronts to solve the technology challenges involved in developing the field. This was not technology in a laboratory, but technology at full scale in the real world, and we learned much from these experiences.

The reservoirs are classed as turbidite reservoirs, the sediments forming them having been laid down as a result of an underwater 'landslide' millions of years ago. Thunder Horse South is a four-way enclosure - that is, enclosed by rocks on all sides - while Thunder Horse North is a three-way, butting up against the salt column on one side, the salt going far deeper into the subsurface than the hydrocarbon reservoirs. Within the reservoirs, hydrocarbons lie in three separate stacked formations, referred to as pink, brown and peach.

'Current production comes mainly from the brown zone,' explains senior reservoir engineer Hong Leung. 'The wells are performing better than expected for two reasons. One is that the reservoir is less compartmentalised than we first thought; the other is the huge scale of the aquifer which lies below the hydrocarbon reservoir, providing the energy to drive the oil to the wells.'

Arnold and his subsurface team are responsible for delivering a robust depletion plan for the reservoirs, a plan that changes as the reservoirs yield more information about their behaviour and the Thunder Horse reservoir model is updated - for example, predicting when water breakthrough may begin to occur in producing wells, or knowing when water injection may be needed to maintain reservoir pressure or sweep out untouched pockets of oil.

'Before we started up the field, we viewed water injection to be an early requirement,' Arnold points out. 'Now that we have more information about the reservoirs, we see water injection more as an optimisation at some later stage.'

'We are very fortunate to have truly world class reservoirs and wells in Thunder Horse, individual wells that can deliver over 50,000 barrels per day,' concludes Arnold. 'Most fields are limited in production terms by their reservoir. Not so for Thunder Horse.'