Fluid properties determine flow line blockage potential

Alex Hunt
Texaco Ltd.
London

A thorough understanding of fluid properties helps in determining the potential of hydrates, paraffins, or asphaltenes to block subsea flow lines.

Thermal, chemical, and mechanical methods are the main ways for preventing deposition. Already in both the North Sea and the Gulf of Mexico, blockages have led to significant losses in production and reserves recovery.

This first article in a two-part series discusses thermal and chemical methods in overcoming fluid behavior problems caused by hydrate and other fluid constituents in subsea multiphase flow.

Subsea production

Subsea technology has advanced markedly in the last 10 years. Much effort has gone into developing suitable hardware, and this is currently continuing as the industry seeks to produce from ever-increasing water depths. However, the industry faces transport challenges as well as the more obvious hardware problems. Two factors are of particular note.

First, most deepwater developments currently have either local floating production systems or subsea tie-backs to shallow water host platforms. Other than in the Gulf of Mexico, where tension leg platforms (TLPs) and Spars seem to be the current "flavors of the month," most development scenarios include subsea trees and multiphase flow lines, possibly with some subsea manifolding as well.

Second, if the subsea system requires subsequent modification after installation, deepwater and no diver intervention place limits on the work that can be performed. This increases costs significantly.

The combination of multiphase flow and costly and limited intervention makes it necessary for process designers to understand in detail the fluid behavior within a subsea system. Fluid behavior may impact both the subsea system design and its operating philosophy and procedures.

Without this understanding, one runs the risk of being forced to abandon flow lines due to blockage. This obviously has a major adverse effect on the development economics.

Possible problems

In general, subsea multiphase systems may have to handle a wide range of potential problems, which may include a combination of any of the following:

• Hydrates
• Paraffins (waxes)
• Resins
• Asphaltenes
• Scales
• Sand
• Emulsions
• Foams
• Slugs
• CO₂
• H₂S.

The operational difficulties that these problems pose are further increased by changes in pressure, temperature, composition, and water cut of the produced fluids as the field depletes. Therefore, at an early stage in the design process, it is necessary to identify problems that may occur as well as the sequence and magnitude of the difficulty that these problems may pose.

Because fields are regularly operated within contractual constraints, such as "send or pay" clauses, subsea multiphase systems often have to be designed for a number of different steady-state conditions at different throughputs set by swing factors and nominations included in transportation and sales contracts. Designs must, therefore, transition smoothly between constraints, often on relatively short notice.

Fields have to be shut down in a controlled manner and be restarted after both planned and unplanned shutdowns of varying duration. Therefore, as early as possible, one needs to establish a matrix for different operating scenarios.

It is worth noting that a recent survey^{1 2} reported that hydrates were considered to be a very significant issue by major oil companies. Curiously, the majors did not consider paraffins to be so important. However, the independent companies had a different view and also ranked paraffins very highly.

This raises the rather interesting question of whether the oil majors are in fact correct or simply less aware of the potential problems of paraffins.

Nucleation, growth, deposition

It is important to note the distinction between nucleation, growth, and deposition.

During nucleation, solid crystals may form, but no operational problems will occur if the crystals are carried along by the flowing fluids.

Growth may occur both before and after deposition, such as at the pipe wall, although difficulties only arise if the solids begin to deposit and stop moving. Again, this problem may be manageable if the deposition is both slow and uniform.

These distinctions are important for the following reason. For a long time, a truism within the oil industry says that hydrates do not cause problems in live oil systems. It has even been stated that hydrates cannot form in oil. This is not true, as recent work by the DeepStar project shows.³

Laboratory work indicates that hydrates can form in all multiphase systems, including both black oils and gas condensates. However, a survey of operators conducted by DeepStar reported that no deposition problems were encountered during normal steady-state operations, even when operating within the hydrate formation envelope. The explanation was that any crystals precipitated are carried along in the flow stream.

Operational difficulties were, however, reported on restart following shutdowns. These were thought to be due to hydrates formed during the shutdown being carried to a point at which the hydrates started to agglomerate, causing blockages.

Thus, precipitation may not cause problems, but to avoid operational difficulties, the possibility of deposition must be assessed and mitigation techniques developed. The DeepStar survey also shows that nothing should be taken for granted in multiphase systems.

Preventing deposition

In general, regardless of whether one considers hydrates, paraffins, or asphaltenes, there are two main methods of preventing deposition, namely thermal and chemical.

Thermal methods

Thermal methods use either the conservation or introduction of heat. In the past, heat conservation has been the normal technique and is simply insulation (Fig. 1 [28372 bytes]).

The thermal limits of high performance insulation materials have been reached, but for deepwater applications, a number of insulation materials are capable of standing up to the external water pressures.^{4 5} For long tie-back distances, in excess of 20 km, insulation is likely to be inadequate.

Alternative possibilities exist, but these have significantly higher capital cost. Some work has looked at ceramic materials replacing conventional insulation, although in the near future it is currently considered that handling and installation difficulties preclude ceramics.

One available option is a pipe-in-pipe configuration. In this case, the outer pipe does not need to be designed for full pressure containment. The annular space holds the insulating material and the outer pipe merely acts as a protective sleeve. The insulating material does not need mechanical strength; therefore, this offers some advantages for very deepwater applications.

Possible insulating materials include glass wool or alumina silicate beads. Also, open cell insulation in an inert gas atmosphere is being examined for higher insulation properties, but this is not yet available commercially.

The 30-km multiphase export pipeline from Texaco North Sea U.K. Co.'s Erskine field uses a bead-filled, pipe-in-pipe system. Gas as an insulator in the annulus is also being considered.

An extension of the pipe-in-pipe concept, proposed for very long tie-back distances, is a vacuum. Here the annulus is evacuated; however, the outer pipe will have to hold the vacuum and major questions remain over the economics and pipe-laying procedures.

A number of different concepts are available for introducing additional heat to a flow line (Fig. 2 [25611 bytes]). The simplest is an external hot-water jacket, either for a pipe-in-pipe system or for a bundle. This is the solution adopted for the Britannia field development operated by Conoco (U.K.) Ltd. and Chevron U.K. Ltd. The heat would be supplied from waste-heat recovery units on the main generators.

Other proposals use either conductive or inductive heat tracing. There is concern over the reliability of conductive systems.

Den norske stats oljeselskap AS (Statoil) is developing one inductive heating system, the Combipipe system.⁶ In this system, three continuous unarmored power cables are inserted in longitudinal grooves in the pipe coating prior to final wrapping.

Once power is supplied, the magnetic field generated around each cable induces electric currents in the pipe wall, thus heating it. The system has been qualified and patented by Statoil and may be installed in an upcoming field development.

A recent development by Petrobras introduces heat by means of an exothermic chemical reaction.⁷ This technique is not applicable for continuous heat input but has been used in a batch process to melt deposits. Known as SGN (Nitrogen Generation System), the process uses two inorganic salts in an organic solvent, plus a chemical activator that initiates the reaction and controls the rate of heat production. The reaction is as follows:

 NH4Cl + NaNO2 N2 + NaCl + 2H2O + DH 

The process successfully has removed substantial paraffin deposits in flow lines and works by a combination of:

• Thermal-due to the heat produced
• Chemical- due to solvent action
• Fluid-mechanical effects-due to flow.

Chemical inhibitors

An alternative to the thermal processes previously described is chemical inhibition. Chemical inhibitors require a much lower capital outlay but may greatly increase operating costs. It is, therefore, important to review the costs over the whole operating life of the field.

At present, inhibitor efficiency, particularly for paraffin and asphaltene mitigation, is difficult to predict, and therefore, the injection volumes required and through-life costs also are hard to assess accurately. Multiphase systems exacerbate the problem and it must be said that the chemical suppliers have only recently become aware of this critical issue.

Well tests provide the fluid samples for initial evaluation and testing. Past experience shows that these samples often do not represent the fluids a field produces over time. Also, the majority of these samples are "dead" and laboratory tests of "live" fluids are normally conducted with synthetic gas fractions and brines.

To avoid potential future operational problems, it is important to note that the chemical injection requirements determined from these tests are conservative. But the industry can no longer afford the costs associated with this conservatism. This applies to all chemical inhibitors.

However, when benchmarked against actual field experience, predictions for hydrate inhibitor requirements made with the results of tests on live laboratory samples have improved markedly over the last few years.

Hydrate predictions

Hydrate formation conditions based on fluid sample compositions are normally predicted with computer subroutines within process simulation packages such as Hysim and Provision. The technique is based on statistical thermodynamics, which uses a predictive algorithm with additional experimental data included to modify or "tune" the mathematical predictions.⁸ The algorithm is normally based on the thermodynamic model developed by Van Der Walls and Plateeuw.

Most commercially available packages use algorithms developed by D.B. Robinson & Associates, who also offer a stand-alone hydrate prediction program, Equiphase. However, much of the empirical data used to calibrate and tune the theoretical predictions for multiphase systems are proprietary.

Much of the original data is believed to be derived from single-phase, wet-gas systems, and the validity of this when applied to multiphase systems is open to question because it is difficult to independently evaluate the accuracy.

Also hard to assess are the pressure and temperature ranges over which the predictions have been validated. It is currently believed that the predictions for high-pressure/high-temperature (HP/HT) systems may not be valid, particularly at high pressures.

An alternative to the Equiphase program is Multiflash, developed by Infochem Computer Services Ltd. Although this approach also is based on statistical thermodynamics for hydrate prediction, it is different than Equiphase in that it determines a phase envelope for all species (solids, liquids, and vapors) that may be present, giving a more rigorous solution.

One aspect that has come to light for improving the predictions is the importance of tuning the equations of state used by the process simulation packages. This was discussed in an article in the Oil & Gas Journal.⁹

Typically, cubic equations of state such as Peng Robinson (PR) or Soave Redlich Kwong (SRK) are used with the input data taken from pressure-volume-temperature (PVT) reports. More accurate predictions can be obtained by modifying the binary interaction parameters and adjusting the pseudocomponent fractions to fit the bubble point data in the PVT report. When this is done, the hydrate predictions match the experimental results more closely.

In general, predictions within 3-8° F. of experimental measurements have been obtained for HP/HT systems.

However, also note that the accuracy of laboratory equipment has improved markedly in the last few years. Labs now are able to distinguish between hydrate types (Fig. 3 [22390 bytes]) and to measure both hydrate formation and dissociation conditions. There can be a significant difference between the two, depending on the subcooling applied and the rate of temperature change with time.

The data used by the computer programs tend to be based on dissociation conditions, but these are also conservative. However, it is important to know the point at which hydrates could start to form, even if no problems are actually experienced in practice.

For the future, the most promising alternative to statistical thermodynamics is a technique based on modeling the kinetics of hydrate crystal nucleation, growth, and agglomeration. The Gas Research Institute (GRI) has been considering this, but the large number of potential variables that would have to be accounted for is perceived as the major problem to overcome for this method to become viable.

At present, experimental work sponsored by GRI at King's College, London, is examining the kinetics of crystal nucleation and growth. From this it may be possible to start developing a new modeling technique.

Multiphase flow

Recent work completed as part of the DeepStar project¹⁰ indicates that certain aspects of multiphase flow may impact hydrate formation conditions. The majority of these factors have not yet been quantified; indeed, one of the primary recommendations of this work is that some of them need more detailed study.

It is important to note that hydrates, paraffins, and asphaltenes must not be treated as separate and individual problems. There is some recent experimental evidence that nucleation of one may trigger nucleation of another. This will be reviewed in more detail later.

As discussed earlier, the gas/oil ratio, such as the presence of hydrocarbon liquids, appears to suppress hydrate formation. The presence of cyclo-alkanes in the hydrocarbon liquid phase appears to depress the hydrate formation temperature, while the presence of asphaltenes seems to have the opposite effect.

Paraffin deposition does not prevent hydrate formation but may reduce the potential for agglomeration, primarily because hydrate crystals are prevented from adhering to the pipe walls. Hydrate crystal agglomeration also seems to be reduced in high viscosity oils.

For a long time, produced water salinity has been known to lower hydrate formation temperature, although only some research quantifies the effects. At present, the recommendation is that no margin is placed on the predictions because the need for a margin will be offset by an allowance for salinity effects.

The flow regime within the system also seems to have some effect and it is this that may be the most important factor. In stratified flow, crystals adhere to the pipe wall at the oil/water interface rather than all over the inner pipe circumference.

In slug flow, crystals tend to be carried through the system without agglomerating, because the higher interfacial shear stress caused by slugs will remove the crystals from the pipe wall. This merits further investigation because different operating strategies may affect hydrate blockage.

Hydrate inhibition

There are two methods by which hydrate formation may be inhibited, namely thermodynamic and kinetic. Thermodynamic inhibitors, such as methanol and the glycols, work by depressing the hydrate formation temperature of the hydrocarbon/water mixture. As the water cut increases, more inhibitor is required (Fig. 4 [23720 bytes]).

Simulation programs can estimate the required amount of inhibitor, but the validity of these predictions may once again be difficult to assess because most are based on gas/water mixtures. The empirical adjustments used to compensate for multiphase behavior are hard to validate.

This is particularly true for systems that have not been previously analyzed, such as multiphase systems where the inhibitor partitions between the different phases are present. In these cases, experimental work is also needed to validate and extend the predictions.

The importance of tuning the equations of state has already been discussed. Currently, to estimate methanol volumes and taking account of partitioning, Texaco uses the Provision process simulation package with the Modified SRK equation of state and adjusted binary interaction parameters. This technique has given good results when compared with field data.

Threshold hydrate inhibitors (THIs) work in much lower concentrations than thermodynamic inhibitors; thus, they offer a significant operating cost decrease. Three broad classes have been developed (Table 1 [11810 bytes]):¹¹

1. Kinetic additives prevent the nucleation of hydrate crystals.
2. Growth modifiers control the rate of growth of hydrate crystals.
3. Slurry additives limit the droplet size available for hydrate formation.

Kinetic additives are relatively new for hydrate control, although their use to control scale is accepted. They operate principally in the water phase and aim to prevent crystal nucleation. Kinetic additives are relatively insensitive to the hydrocarbon phase and may therefore turn out to be applicable to a wide range of hydrocarbon systems.

Growth and slurry additives recognize that hydrate crystals may form, but their aim is to prevent agglomeration and hence potential blockages.

Growth modifiers operate at the crystal surface, while slurry additives operate by emulsifying water into the liquid hydrocarbon phase. This limits the water droplet size and hence, the ability of the water phase to coalesce and for crystals to agglomerate and form a hydrate block. Slurry additives are therefore able to operate inside the hydrate formation conditions.

Growth modifiers may also form a hydrate phase dispersed in the multiphase system. The nature and behavior of the dispersed phase depends on the additive used to control crystal growth.

Additives that operate by emulsification of water into the hydrocarbon liquid phase are expected to be relatively specific to the hydrocarbon and may therefore need to be tailored to specific field compositions. They are not expected to be suitable for hydrate control in dry gas systems.

Successful field trials of kinetic additives have been conducted and they are now ready for operational deployment.¹² However, growth modifiers and slurry additives are not as well advanced at present and are currently undergoing flow loop tests as part of different joint industry projects. RF-Rogaland Research is conducting a recently initiated project looking primarily at growth modifiers.¹³

Project Eucharis, being conducted by BP Research and Institut Franais du Petrole (IFP), administered by the Petroleum Science & Technology Institute (PSTI) and sponsored by BP Exploration Operating Co. Ltd., British Gas Exploration & Production Ltd., Conoco (U.K.) Ltd., Elf Exploration PLC, Texaco Britain Ltd., Total Oil Marine, and the offshore safety division of the U.K. Health & Safety Executive, is currently testing different types of low dosage additives developed by BP and IFP. The last part of the project is currently under way, with the additives being tested in the 2-in. flow loop at IFP's facility in Solaize.

References

1. Moritis, G., "Study ranks oil, gas industry technology needs," OGJ, Oct. 30, 1995.

2. Moritis, G., "Impact of production and development RD&D ranked," OGJ, Oct. 30, 1995.

3. Operational Experience with Hydrate Formation in Liquid Hydrocarbon Lines, DeepStar II Project Report No. DSII CTR 230-1, June 1994.

4. Oram, R.K., "Advances in Deepwater Pipeline Insulation Techniques and Materials," Deepwater Pipeline Technology Congress 1995, London, Dec. 11-12, 1995.

5. Insulated Flowline Option-Identification Study, DeepStar IIA Project Report No. DSIIA CTR A601-a, August 1995.

6. Kirkhorn, S.S., "Developments in Hydrate Control Employed," Advances in Multiphase Operations Offshore Conference, London, Nov. 29-30, 1995.

7. Khalil, C.N., "New Process for the Chemical De-Waxing of Pipelines," Advances in Multiphase Operations Offshore Conference, London, Nov. 29-30, 1995.

8. A Critical Review of Hydrate Formation Phenomena, Health and Safety Executive Offshore Technology Report No. OTH 93 413, 1994.

9. Dharmadhikari, S., "PVT data useful for design of oil production facilities," OGJ, May 22, 1995.

10. Hydrate Formation Tests-requirements for Multiphase Oil Systems, DeepStar IIA Project Report No. DSIIA CTR A209-1, August 1995.

11. Goodwin, S., and Hunt, A.P., "Prediction, Modeling and Management of Hydrates Using Low Dosage Additives, Advances in Multiphase Operations Offshore Conference, London, Nov. 29-30, 1995.

12. Corrigan, A., et al., "Trials of Threshold Hydrate Inhibitors in the Ravenspurn to Cleeton Line," Paper No. SPE 30696, SPE Annual Technical Conference, Dallas, Oct. 25, 1995.

13. Kelland, M.A., et al., "A New Generation of Gas Hydrate Inhibitors," Paper No. SPE 30696, SPE Annual Technical Conference, Dallas, Oct. 25, 1995.

Based on a presentation at the Advances in Subsea Technology Conference, Aberdeen, Apr. 23-24.

The Author