Norwegian Testing of Emulsion Properties at Sea––The Importance of Oil Type and Release Conditions

https://doi.org/10.1016/S1353-2561(03)00016-1Get rights and content

Abstract

This paper is a review of the major findings from laboratory studies and field trials conducted in Norway in recent years on the emulsification of oils spilled at sea. Controlled bench-scale and meso-scale basin experiments using a wide spectrum of oils have revealed that both the physico-chemical properties of the oils and the release conditions are fundamental determinants of the rate of emulsion formation, for the rheological properties of the emulsion formed and for the rate of natural dispersion at sea.

During the last decade, several series of full-scale field trials with experimental releases of various crude oils have been undertaken in the North Sea and the Norwegian Sea. These have involved both sea surface releases, underwater pipeline leak simulations (release of oil under low pressure and no gas) and underwater blowout simulations (pressurized oil with gas) from 100 and 850 m depth. The field trials have been performed in co-operation with NOFO (Norwegian Clean Seas Association for Operating Companies), individual oil companies, the Norwegian Pollution Control Authority (SFT) and Minerals Management Services (MMS). SINTEF has been responsible for the scientific design and monitoring during these field experiments. The main objectives of the trials have been to study the behaviour of different crude oils spilled under various conditions and to identify the operational and logistical factors associated with different countermeasure techniques.

The paper gives examples of data obtained on the emulsification of spilled oil during these field experiments. The empirical data generated from the experimental field trials have been invaluable for the validation and development of numerical models at SINTEF for predicting the spreading, weathering and behaviour of oil released under various conditions. These models are extensively used in contingency planning and contingency analysis of spill scenarios and as operational tools during spill situations and combat operations.

Introduction

Prior knowledge of the likely behaviour of a spilled oil and pre-spill analyses of the feasibility of different response strategies under various environmental and release conditions should be an essential part of any oil spill contingency planning. Predicting the amount of damage that might occur in various oil spill scenarios enables the effectiveness of alternative response strategies to be assessed. Such a pre-spill analysis has been formalized as the NEBA process (“Net Environmental Benefit Analysis”) of a combat operation (e.g. Baker, 1995, Baker, 1997).

In Norway, the responsible party takes the lead in responding to an oil spill. This is in accordance with the “principle of internal control” that is used by the pollution authorities. The Norwegian Pollution Control Authority requires well-documented contingency plans for refineries, oil terminals and offshore installations. The SINTEF Oil Weathering Model (OWM, e.g. Daling et al., 1997, Daling and Strøm, 1999) and the OSCAR (Oil Spill Contingency and Response) model system (Aamo et al., 1996, Aamo et al., 1997a, Aamo et al., 1997b, Reed et al., 1995, Reed et al., 1996, Reed et al., 1997, Reed, 2001) is now extensively used in such contingency planning. Scenario analysis is used to quantify the fate, weathering, potential environmental consequences and the feasibility and effectiveness of various mitigation methods.

The behaviour of spilled crude oils and refined oil products depends on:

  • the release conditions (the rate and amount of oil spilled, surface release or underwater release, presence of gas, etc.);

  • the prevailing environmental conditions (e.g. temperature, sea-state, currents);

  • the physico-chemical properties of the spilled oil and its propensity to disperse into the water column or to form stable water-in-oil (w/o) emulsions on the sea surface.


For example, the Gullfaks crude spilled at the Braer incident in the Shetlands had a very low content of waxes and asphaltenes, which are important compounds for stabilizing w/o emulsions formed on the sea surface. This, combined with the exceptionally violent weather conditions that prevailed at the time of the spill, resulted in almost all of the 84,000 tons of the spilled Gullfaks crude oil being naturally dispersed (Ritchie et al., 1993).

During the Sea Empress spill of Forties Blend crude oil (another North Sea crude oil) a significant amount of the surface oil was converted into w/o emulsions (Lunel et al., 1996). The feasibility of various countermeasure techniques such as chemical treatment, burning or mechanical recovery, would therefore be greatly influenced both by the release conditions and changing properties exhibited by the weathered oil residue or the w/o emulsion that have been formed.

Very heavy fuel oils like the industrial heavy fuels oils spilled by the Erika and the Baltic Carrier form water-in-oil emulsions slowly. However, many spilled crude oils will rapidly form w/o emulsions when spilled at sea (e.g. Lewis et al., 1995a, Lewis et al., 1995b). Such emulsions will initially have low viscosities, will be unstable, and will tend to revert to the oil residue and water if they are removed from the mixing action of the sea. Unstable emulsions are simply mixtures of water droplets in oil and the w/o emulsion present on the surface will be the result of the dynamic equilibrium of emulsion formation and emulsion breakdown (see Fig. 1(a)). As the viscosity of the oil residue increases due to the evaporative loss of more volatile components and the precipitation of stabilizing agents (asphaltenes, photo-oxidized compounds (resins) and in some crude oils precipitated waxes) the emulsion becomes more stable. The precipitated asphaltenes create an elastic “skin” between the water droplets and the oil (see Fig. 1(b)). The stability of the emulsion will increase because the water droplets cannot coalesce and drain so easily from the emulsion and the equilibrium will tend to favour emulsion formation. After an extended period of weathering and mixing at sea the w/o emulsion produced will have a very high viscosity and be very persistent.

In the early stages of a spill, the rate of natural dispersion will depend [in a large measure] on the sea state. Breaking waves provide the energy needed to disrupt low viscosity (non-emulsified) oils and produce the very small oil droplets with diameters below 50–100 μm that may be considered to be permanently dispersed (Delvigne & Sweeney, 1988). One of the factors that resist the process of natural dispersion is the mechanical strength of the oil or w/o emulsion layer. Weathered oils (Berger et al., 1993) and w/o emulsions (Sherman, 1968) are liquids or semisolids with more complex rheological characteristics and will exhibit high apparent viscosities, an elastic component and/or a definite yield stress. W/o emulsification therefore retards the rate of natural dispersion. Natural dispersion and w/o emulsification are therefore “competing processes” as illustrated schematically in Fig. 1(a).

Thick layers of oil will be able to accommodate entrained water droplets of greater size more easily than thin film oil layers. Very thin oil layers will be disrupted by the presence of water droplets and will split to release the water. The thickness of the oil layer therefore has an effect on the ease and rate of w/o emulsification. The relative rates of natural dispersion and w/o emulsification will therefore depend on the initial film thickness (and this will be determined by release conditions), the sea state (including the presence or absence of convergence zones) and the physico-chemical properties of the oil. At low or intermediate sea states the rate of natural dispersion will initially be high but will be reduced and eventually cease as the oil remaining on the sea surface becomes more viscous due to evaporation of lighter components and increased viscosity due to the formation of a w/o emulsion (Lewis et al., 1994).

The turbulence provided by breaking waves is locally intense, but its effect decreases with weathering of oil. If the apparent viscosity of the oil or w/o emulsion is very high or if it possess a significant elastic component, the prevailing turbulence may be unable to break up the slick into oil droplets. There will therefore be no significant amount of droplet formation or natural dispersion. At higher sea states, the turbulence will be able to break up viscous emulsions into fragments which may appear as large lumps which will not naturally disperse but may be washed-over by waves for extended periods. While not being naturally dispersed, this mechanism will apparently remove the bulk of oil pollution from localized areas by distributing it across much larger areas of the sea surface. This is illustrated schematically in Fig. 2. The life-time on the sea surface of emulsified oil slicks and fragments of surface slicks will therefore be relatively longer (under otherwise similar sea conditions) than a light, non-viscous, non-emulsifying oil (like a condensate or a diesel oil).

Section snippets

Emulsification Studies in Laboratory and Weathering Flume Basins

At SINTEF there has been a continuous R&D activity on oil spill weathering and behaviour at sea since the Ekofisk Bravo accident in 1977. The SINTEF approach for characterization and predictions of the oil weathering properties has involved:

  • Bench-scale laboratory weathering studies;

  • Meso-scale flume basin weathering;

  • Oil Weathering Model (OWM);

  • Verification and correlation of laboratory data and model predictions through “ground-truth” data from experimental field trials.


A standardized laboratory

Full-Scale Field Trials in Norway

Unique field data have been obtained on emulsification and behaviour of oils spilled during four series of experimental releases with various oils simulating various release conditions. These studies involved both surface releases (1994, 1995, and 1996), simulated sub-sea pipeline leaks (release of oil and no gas, 1994), and simulated sub-sea blowouts (oil with gas) released both from 100 m depth (1996) and 850 m depth (DeepSpill June 2000). Three of the field series were carried out in the

Conclusions and Recommendations

Laboratory experiments and full scale field experiments have been performed in Norway during the last decades in order to obtain more knowledge about the emulsification of crude oils. A broad range of oils has been investigated in the laboratory both in bench-scale and meso-scale. Field trials have been performed to verify laboratory results, for calibration of models and to study the influence of different release conditions. The major findings from the work are as follows:

Controlled

Acknowledgements

The Norwegian Clean Seas Association for Operating Companies (NOFO) is acknowledged for supporting this review paper.

References (36)

  • P.S. Daling et al.

    Characterization of crude oils for environmental purposes

    Oil and Chemical Pollution

    (1990)
  • P.S. Daling et al.

    Weathering of oil at sea; model/field data comparisons

    Spill Science and Technology Bulletin

    (1999)
  • G.A.L. Delvigne et al.

    Natural dispersion of oil

    Oil and Chemical Pollution

    (1988)
  • H. Rye et al.

    Subsurface blowouts: results from field experiments

    Spill Science and Technology Bulletin

    (1997)
  • Aamo, O.M., Downing, K., Reed, M., 1996. Calibration, verification, and sensitivity analysis of the IKU oil spill...
  • Aamo, O.M., Reed, M., Downing, K., 1997a. Oil spill contingency and response (OSCAR) model system: sensitivity studies....
  • O.M. Aamo et al.

    Regional contingency planning using the OSCAR oil spill contingency and response model

  • J.M. Baker

    Net environmental benefit analysis for oil spill response

  • Baker, J.M., 1997. Differences in risk perception: How clean I clean? Issue Paper at the 1995 Oil Spill Conference....
  • Berger, D., Mackay, D., Charles, M.E., Ooijen, H.V., 1993. Rheological characterisation of waxy crude oils for...
  • Brandvik, P.J., Lewis, A., Strøm-Kristiansen, T., Hokstad, J.N., Daling., P.S., 1996. NOFO 1996. Oil on water...
  • Daling, P.S., Aamo, O.M., Lewis, A., Strøm-Kristiansen, T., 1997. SINTEF Oil Weathering Model––predicting oil’s...
  • Fanneløp et al.

    Hydrodynamics of underwater blowouts

    Norwegian Maritime Research

    (1980)
  • Hokstad, J.N., Daling, P.S., Lewis, A., Strøm-Kristiansen, T., 1993. Methodology for testing water-in-oil emulsions and...
  • Johansen, Ø., 1991. Numerical modeling of physical properties of weathered North Sea crude oils. DIWO-Report No. 15....
  • Johansen, Ø., Rye, H., Melbye, A.G., Jensen, H.V., Serigstad, B., Knutsen, T., 2001. DeepSpill JIP––experimental...
  • A. Lewis et al.

    Chemical dispersion of oil and water-in-oil emulsions

  • Lewis, A., Daling, P.S., Strøm-Kristiansen, T., Brandvik, P.J., 1995a. The behaviour of Sture Blend crude oil spilled...
  • Cited by (52)

    • A brief survey of oil spill weathering models

      2021, Marine Hydrocarbon Spill Assessments: From Baseline Information through to Decision Support Tools
    • Small-scale in-situ burning (ISB) experiments with chemically confined crude oils on water

      2020, Fire Safety Journal
      Citation Excerpt :

      The artificial water-in-oil emulsions were made for both crudes by using a modified version of the rotatory flask technique and had a water content of 25% [27]. The details of this technique can be found in Ref. [18]. It should be highlighted that this technique is not carried out according to ASTM F3045, which is a standard for preparing and classifying water-in-oil emulsions (that is not part of the current study's objectives).

    View all citing articles on Scopus
    View full text