Long-range energy forecasts suggest world demand for LNG could double by 2020. This robust demand trend has driven the diversification of LNG concepts, leading in the near-term deployment of the first LNG floating production storage and offloading unit.
Industry efforts at moving FLNG concepts to market readiness have been directed at qualifying technology for LNG liquefaction, feed-gas treatment, product containment and offloading systems.
The resilience of these systems to the metocean characteristics relevant to the development province has been the subject of rigorous examination.
There is growing confidence in the industry that ship and topsides technology has been adequately adapted to marine requirements.
However, still under-explored are the issues arising from marrying the subsea architecture (supplying feed-gas) with the processing topsides of the LNG FPSO.
This paper investigates the concept critical and operational challenges arising from marrying subsea architecture with the LNG FPSO, and assesses possible solutions.
Can the subsea architecture be modified to mitigate slugging? In the extreme, is a slug handling/NGLs removal unit required upstream of the LNG FPSO?
Compatible with flow assurance, what design safeguards can be introduced subsea and topsides to assure the performance integrity of the topsides absorption systems?
Figure 1 illustrates the deployment of the LNG FPSO which receives its feedgas from subsea wells in the non-associated gas field.
Flowlines connect the subsea wells to the FPSO, with risers conveying the gas through a swivel located in an external turret.
LNG and any product LPG/condensate stripped out from the feed-gas is stored and periodically shipped by trade carriers to destination markets.
In FLNG applications, the FPSO is deployed in the vicinity of the subsea architecture unlike for an onshore LNG plant, where the feed-gas pipeline transports gas from a relatively remote gas source.
There is little opportunity for line-pack to cushion supply-side volatilities. The topsides system is also significantly exposed to transient operations (turn-up, turn-down, etc.), characteristic of subsea system operation.
The quality of gas (e.g. condensate to gas ratio), the subsea architecture and flowline sizes will determine wet-gas flow stability and slugging propensity of the system. Unlike for the onshore LNG plant, the opportunity for substantial slug capture on the floater topsides is obviously limited.
The slugging in the risers can lead to pulsating flows, which can have a detrimental effect on the stability of the topsides pre-processing, NGLs fractionation and liquefaction operations.
The flow-assurance measures adopted for subsea gas production represent a further area of design concern.
The injection of chemicals to suppress hydrates, mitigate wax formation and inhibit corrosion
can potentially result in the ingress of these chemicals into the pre-processing systems, including the Acid Gas Removal Unit Absorber.
These chemicals could promote foaming and degradation of the solvents which are a potential threat to plant availability and performance.
Figure 2: The FPSO topsides flow diagram for processing units in the FLNG concept of feed-gas preparation following by liquefaction to make LNG
The extraction of LPG and condensate is to a greater and lesser extent integrated with the pre-cooling steps associated with liquefaction step, depending on the generic configuration of the process.
The tighter the integration, the more efficient is the LNG scheme. The principal facilities on the FPSO include:
- Inlet Gas Reception Facilities
- Acid Gas Removal Unit (AGRU)
- Gas Dehydration
- Mercury Removal Unit
- Gas Conditioning Unit
- Gas Liquefaction Unit
- LPG/Condensate Fractionation
- Product Storage & Offloading
- Utility Systems
- Safety Systems
The topsides systems are located on a plated deck platform approximately 3.5 metres above the vessel deck.
The storage tanks for LNG, LPG and condensate are located below deck. The living quarters are located close to the bow of the FPSO upwind of the hazardous areas, flare etc.
The core of the processing is the LNG liquefaction block for which there are several candidate technologies, most of which are proprietary offerings. They include these technology categories:
- Single Mixed Refrigerant
- Dual Mixed Refrigerant
- Propane Pre-Cooled Mixed Refrigerant
- Turbo-expander based cycles
Details of these processes are available from several sources in the public domain. Prospective floating liquefaction units appear to be polarising into two capacity bands with different technology selection implications.
Shell and Petrobras have front-end engineering and design concluded on the potential application of LNG FPSOs for capacities of 3 million tonnes per annum or more, which are close to baseload plant Train capacities, and are consistent with the exploitation of gas reserves of 4-5 trillion cubic feet and greater.
On the other hand, companies such as FLEX LNG and Hoegh LNG are offering capacities in the sub-2 MTPA capacity domain, principally targeted at the exploitation of gas reserves of 0.5 to 3 Tcf, more characteristic of the mid-tier stranded gas reserves.
Technology evaluations carried out in previous studies by the author have demonstrated that ‘best fit’ liquefaction technology differs for the 2-plus MTPA capacity range and the sub-2 MTPA range.
- For example, the Cascade process, and several Mixed Refrigerant processes have been shown to be an attractive technology options when considering applications in the 2-plus MTPA bracket
- In the small-scale category the Dual Turbo-Expander Cycles (e.g. BHP/Kanfa Aragon/CB&I) and the SMR processes appear to fit well in the sub-2 MTPA capacity size, and are suitable for application to the mid-tier reserves
- Shell’s LNG FPSO concept, for a capacity of 3.5 MTPA is premised on the application of the dual mixed refrigerant technology, thereby securing the higher efficiencies, and lower specific power .consumption essential for upscale processing.
The Flex LNG offering is based on the application of Kanfa Aragon’s dual nitrogen turbo-expander cycle.
Although the turbo-expander cycles have a penalty to pay in terms of specific power consumption, there are several important advantages for the latter concept when it comes to floating production application, and particularly if there is variability in the gas-feed rate and composition.
In the evaluation of LNG FPSO technology, the principal focus of discussion has hitherto been on the choice of the liquefaction technology, the front-end processing (i.e. the gas treatment and NGL/condensate stabilisation), and marine usage issues.
Few assessments are available on the integration aspects of the subsea system that will provide the gas feedstock to the LNG FPSO unit.
In this section, I will look at the subsea architecture upstream of the LNG FPSO, and the key issues that must be managed both in design and operations to ensure that the subsea and topsides systems are seamlessly integrated.
The function of the subsea architecture is to gather the gas from the wellheads and route this to the LNG FPSO.
As noted earlier, LNG FPSO capacities can vary considerably. Provision of capacity of 1.5 MTPA will require a feed-gas rate of approximately 250 MMscfd.
The 4.5 MTPA unit, now being considered for Inpex’s Abadi FLNG development will require up to 18 wells and a gas-feed rate estimated at approximately 750 MMscfd.
It is clear that the subsea architecture will differ considerably between the smaller capacity units and the larger units envisaged by Shell, Petrobras, and Inpex, which will require a far ore complex network.
Feed-gas rate and LNG capacity are not the only variables. The composition of the gas (as produced from the subsea wells) can exhibit significant variations.
Gas fields can range widely on condensate-to-gas ratios (CGRs). CGRs can typically vary from 1-2 bbls/MMscf to 20 + bbbls/MMscf.
Likewise, the CO2 content of the gas can also vary widely. In recognition of this, the Shell LNG FPSO is being designed for a maximum CGR of up to 60 bbls/MMscf, and CO2 content of up to 13 percent.
Depending on the extent of the reservoir, the required complexity of the subsea gas gathering architecture can vary between developments.
A large areal extent combined with a high gas demand can result in a gas-gathering network that is complex to analyse, to design and ultimately to operate.
The architecture of the gathering system is developed on the basis of the location of the wells and the LNG FPSO, and modelled utilising one of the proprietary multiphase analysis software suites such as OLGA.
Evaluation of the gathering system requires a thermo-hydraulic analysis of multiphase flow from the wellheads through the gathering system to the LNG FPSO.
This evaluation, initially undertaken in the steady state mode, parametrically analyses pipe size selection across several alternative gathering system configurations, and terrain geometry.
This process eventually leads to the choice of the optimum network configuration that best balances hydraulic performance and considerations of flow assurance.
Figure 3 illustrates the effects of line size on the backpressure at the gas field
Multiphase flow analysis of the gathering system network yields a number of results which have an impact on the design of the topsides of the LNG FPSO.
The objective of this multiphase hydraulic modelling is to arrive at sizing of the gathering system pipeline segments which will enable the wet gas to be transported to the LNG FPSO, and arrive there at the desired pressure - while simultaneously ensuring that the back pressure at the wells is at or below the desired depletion pressure at the wells.
The selection of line size must be optimised against gas production from the well as wellhead pressures decline through field maturity.
However, several other factors will also drive this decision, and these are discussed next.
The propensity to slug for a given pipeline will be determined by the two-phase flow regime in the flowline system, which in turn is determined by flow rate, CGR, terrain geometry etc. as previously discussed.
These flow regimes encompass a variety of flow patterns which include the following:
- Annular Dispersed/Mist
- Stratified Wavy
- Slug (intermittent)
- Dispersed bubble.
Flow patterns for a given flowline configuration and size will vary with the turn-up and turn-down of flow in the pipeline. A principal concern with the operation of the topsides system is the resilience of the operations to slugging.
Let us look at the role of liquid hold-up in the generation of slugs.
Liquid hold-up represents the amount of liquid resident in the flowline at a particular gas rate.
As flowrate is reduced during turndown, the equilibrium liquid hold-up increases. Thus after a turndown period, ramp-up of production can lead to re-adjustment of the liquid hold-up to a lower equilibrium level.
This can lead to a surge of liquid reaching the topsides, potentially overwhelming the topsides slug catcher/separator.
There are other mechanisms too for the formation and propagation of slugs. Significant changes in the terrain profile can also lead to slug formation and propagation.
Restart of operations after a shutdown can also result in travel of slug to the topsides.
Dynamic analysis is often employed, after steady state analysis is concluded, to analyse transient operations and to test the robustness of the selected architecture.
This type of analysis is employed to evaluate turn-down and ramp-up scenarios, and to determine gas gathering system behaviour including cool-down impact of extended shutdown, depressurisation, and cold restart.
It is important to design and size the flowline system to enable operations to be conducted above the critical gas rate, and also to manage ramp-up operations carefully.
For a given size of slug catcher, the rate of ramp-up should be managed to ensure that the liquid arrival does not induce pulsating gas flow to the AGRU and liquefaction units.
The preceding analysis has identified slugging as a key interface issue between the subsea and topsides.
What are the possible solutions for managing slugging, and ensuring the topsides systems will see as stable a flow regime as possible?
The performance of Absorber systems in the AGRU and the liquefaction circuits require steady gas flow rates for optimal and on-specification performance.
The avoidance of pulsating gas flow through these systems is essential to ensure stable
performance and avoid operational upsets.
The design of an adequately sized slug catcher is clearly a pre-requisite. However, unlike in an onshore plant, there are limitations to what slug catcher size can be selected, given the limited plot space on the vessel.
Whereas for onshore plants, pipe finger type slug catchers with capacities of 2000 cubic metres and beyond have been installed, the slug catcher that can be realistically deployed on an FPSO will have a maximum capacity of a few hundred bbls only.
Given this constraint, it becomes imperative to control both the formation of the slug, and the rate of receipt into the slug catcher to ensure that production operations are not impacted.
Several options exist for the mitigation of slugs and these are listed:
- Through careful optimisation of gathering system design ensure that stable flow range is maximized, and hold-up is minimized through selection of lowest diameter feasible (without exceeding the erosional velocity limits of the flowlines)
- For a given bathymetry of the sea floor, select the location of the FPSO if possible such that the flowline runs up slope to the base of the riser. This will minimise the size of the slug and the propensity to slug in the riser.
- Consider some form of predictive control that will sense the onset of slugs and activate a choke at the inlet to the slug catcher. This will have the effect of controlling the rate of receipt of the slug, thus ensuring that the slug can be contained (within a specified size of slug catcher), and pulsation in flow is minimised. However, the reduction in gas rate into the AGRU may be unavoidable.
Consider the application of gas lift to the base of the riser. This will ensure that no undue build-up of slug will occur at the riser base, as the gas lift will enable a more uniform two phase transport of fluids up the riser.
In conjunction with the above design steps, several operational procedures need to be implemented to manage multiphase operations, and mitigate slugging and disturbance to operations of the AGRU, the fractionation system and the liquefaction circuits.
- Manage subsea flow rates to stay above the critical gas rates as determined by the hydraulic simulations
- Ensure ramp-up operations and ramp-up rates are managed in accordance with the operating envelopes determined by the dynamic analysis carried out
- Develop operating procedures for subsea system restart that are compatible and integrated with restart operations of the pre liquefaction and liquefaction processing steps
In the extreme, high CGR gas-feed and severity of slugging could also require consideration of more radical solutions.
Large-size slug catchers, associated liquids handling systems, and LPG/condensate fractionation could be located on a separate floater, i.e. give consideration to installation of an LPG FPSO in tandem with the LNG FPSO.
The economics of this more radical alternative, expected to be marginal, will be determined by volumes of LPG and condensate that can be extracted from the rich gas feedstock.
Another less conventional alternative is to consider the installation of a subsea slug catcher in proximity of the LNG FPSO, along the lines deployed on the Highlander installation in the North Sea.
In this project, a subsea slug catcher module is deployed on the seabed adjacent to the base of the jacket. The gas and liquid are separately routed to the topsides (the liquids pumped by downhole pumps deployed in risers installed in the platform).
A parallel concept can be considered for the LNG FPSO, with seabed located pumps delivering the liquids to the topsides.
The management of hydrates in the subsea system has significant ramifications for the LNG FPSO topsides.
Based on flow assurance studies, the preferred hydrate suppression regime has to be configured.
The most common hydrate inhibitors used are methanol and glycol, which act thermo- dynamically to shift the hydrate dissociation curve such that the operating temperature falls outside the hydrate region.
Depending on the differences between the operating temperature and the hydrate formation temperature, hydrate management can also be accomplished via the injection of low dosage hydrate inhibitors (LDHI).
These LDHIs are either Kinetic Hydrate Inhibitors (KHI) or Anti-Agglomerants (AA), the former relying on delaying of hydrate nucleation and crystal growth processes, and the latter on dispersing the hydrates into the liquid hydrocarbon phase, and keeping the crystals in a slurry form.
If a MEG system is adopted, then depending on the concentration of glycol required, and levels of saturated and formation water in the gas, there could be a significant traffic of MEG that is received back at the platform for regeneration, reclamation and storage.
This could represent a significant demand on topsides weight and space. Since the MEG after going through the slug catcher/inlet separator will be removed in the water phase, there is less concern with MEG carryover into the gas phase impacting solvent operations in the AGRU unit.
If methanol is employed for hydrate inhibition, we would anticipate carryover of methanol with the gas phase, with potential ingress into the Absorber system.
In the event of an extended shutdown, the cool down of the flowline system can result in the gas entering the hydrate formation region.
It will become necessary to depressure the flowline network, to ensure pressure/temperature is outside of the hydrate formation region.
The flare system on the FPSO will need to be checked out to ensure that the depressurization of the flowline network can be accomplished, well before the cool down of the system can occur to the hydrate formation region.
Foaming in the Amine sweetening process is an often reported problem, and can be caused by a number of reasons, which include:
- Suspended solids, including pipeline corrosion products/ fines
- Hydrocarbon liquid carryover
- Methanol build-up
- Ingress of corrosion inhibitors
An often cited reason for foaming in Amine Absorbers is the presence of liquid hydrocarbons.
It is essential to ensure, in the design of the pre-processing circuits, that liquid carryover either through the effects of slugging or through condensation does not result in the ingress of liquid hydrocarbons.
It should also be noted that a cocktail of chemicals can be injected at the wells, including corrosion inhibitors, wax inhibitors etc. Carryover of these and solids (rust, scale, and sand fines from the flowlines) can exacerbate foaming problems.
I have highlighted the areas which merit attention in marrying subsea architecture with processing on the topsides of the LNG FPSO.
I have also identified the critical flow assurance issues which can have an impact on the design and operation of the LNG FPSO. These include the generation, propagation and travel of slugs to the topsides.
Key recommendations underscore the need for configuring of the subsea architecture to minimize liquid holdup and slugging potential, and then mitigating and managing slugs when received on the platform.
The measures needed for minimising pulsating flow to the AGRU and liquefaction units are also outlined.
I have also assessed the requirements of hydrate management systems on the topsides, and the implications on topsides if hydrate management requires the deployment of MEG regeneration and reclamation systems.
The importance of protecting the Amine Solvent in the AGRU Absorbers against the carryover of methanol, corrosion and wax inhibitors, corrosion products, sand and other fines is also stressed to ensure a high availability for the topsides systems.
Subsea architectures which are areally dispersed, and consist of a large number of wells, such as may be required for the higher capacity LNG FPSOs, lead to complex gathering networks.
The flow assurance implications of such systems are more profound, and require critical focus and analysis to deliver a seamlessly integrated gas gathering and LNG production facility.
Joe T. Verghese, WorleyParsons Europe Ltd.
This article is based on extracts from a paper entitled: “The seamless integration of Subsea Architecture with Floating LNG - Issues, Challenges and Solutions” by Joe T. Verghese, WorleyParsons Europe Ltd.
1 Gilmour N., Shell Upstream International “Floating LNG: Shell’s Recent History and Current Approach” LNG 16, April 2010
2 Verghese J.T. Monetising the Smaller Gas Reserves - Niche LNG Technology for Mid-Scale LNG Applications”, GasTech 2005, March 2005.
3 Kitamura T.,Inpex Corporation “Outline of Inpex’s Ichthys and Abadi Projects” LNG 16, April 2010
4 Torpe H, et al., Statoil, “Liquid Surge Handling at Asgard by model predictive control”, Technical Conference (Multiphase Pipeline Issues), November 2009.
5.Harun. A.F. et al., “Assuring Gas Deliverability from BP First Subsea Tie-back to Shore”, Technical Conference (Multiphase Pipeline Issues), November 2009.