17
where sg = specific gravity at stock tank conditions (1 bar, 16 C). The number constants are selected so that pure water has an API of 10. Oils with API less than 10
would be non-buoyant in freshwater but might be slightly buoyant in seawater. While
spilled oil density usually increases with weathering, it is rare that the resulting density change will cause a buoyant oil to sink. Instead, other processes may interact to
cause submergence. During DWH, aggregation with marine snow (Passow and
Ziervogel 2016) may have played a major role in causing oil to settle on the bottom.
Industry refers to oil that does not contain dissolved gases as “dead oil.” Such
oils show relatively little change to density as pressure increases. However, deep
well-released oils such as that from DWH are “live oils” and contain significant
amounts of dissolved gases, mainly methane, ethane, propane, and butane. Some
dissolved gases will escape during the pressure drop from reservoir to water surface.
The oil formation volume factor calculates the change in oil volume from reservoir
conditions to the resulting volume of liquid and gas if it were directly brought to
stock tank conditions. For DWH, the oil formation volume factor was about two and
a third (Hsieh 2010). This does not represent the actual observed change between
reservoir volume and surface spill volume since dissolved gas release, other weathering processes, and incorporation of surrounding seawater are not included.
Like most fluids, the density of the oil increases as the temperature decreases.
The increase parameter is a nonlinear function of temperature and density (ASTM
2007) but can be approximated as linear over conditions outside the reservoir.
Another important bulk property is viscosity. Unfortunately, the term relates to
two different properties with different dimensional units. Kinematic viscosity has
dimensions of area/time with its SI unit being the stoke. Dynamic viscosity, sometimes called absolute viscosity, is kinematic density multiplied by the oil density. Its
SI unit is the poise. The oil industry traditionally used neither but instead would use
the time for an oil sample to flow through a certain type of measuring viscometer
(e.g., Saybolt universal second). Fortunately, this is less common today, and most
large oil property libraries store viscosity in one hundredth of poise or centipoise.
Oil viscosity is highly sensitive to temperature change. Past practice in older
surface spill models was to utilize the Eyring’s equation (1935) to calculate (kinematic) viscosity
v v
k T T
v
0
0
1
1
=
−
ref
ref
exp
(2.2)
at the environmental temperature T 0 (K) by extrapolating from some measured laboratory viscosity v ref and temperature T ref . Choice of k v varied depending on the model,
but a typical value, expressed in Kelvin, would be 5000 K (Bobra and Callaghan
1990). Industry itself needed more accurate estimates over greater environmental
extremes, so more complex methods have been developed (e.g., Orbey and Sandler
1993; Abu-Eishah 1999; Hemmati-Sarapardeh et al. 2013) but are not necessarily
imported into current spill modeling.
While mass loss through evaporation (for surface oil) or dissolution will also increase
viscosity, the most significant cause of viscosity increase for susceptible oils is water-inoil emulsification where the emulsified oil viscosity can increase by more than an order
2 The Importance of Understanding Fundamental Physics and Chemistry of Deep Oil…
where sg = specific gravity at stock tank conditions (1 bar, 16 C). The number constants are selected so that pure water has an API of 10. Oils with API less than 10
would be non-buoyant in freshwater but might be slightly buoyant in seawater. While
spilled oil density usually increases with weathering, it is rare that the resulting density change will cause a buoyant oil to sink. Instead, other processes may interact to
cause submergence. During DWH, aggregation with marine snow (Passow and
Ziervogel 2016) may have played a major role in causing oil to settle on the bottom.
Industry refers to oil that does not contain dissolved gases as “dead oil.” Such
oils show relatively little change to density as pressure increases. However, deep
well-released oils such as that from DWH are “live oils” and contain significant
amounts of dissolved gases, mainly methane, ethane, propane, and butane. Some
dissolved gases will escape during the pressure drop from reservoir to water surface.
The oil formation volume factor calculates the change in oil volume from reservoir
conditions to the resulting volume of liquid and gas if it were directly brought to
stock tank conditions. For DWH, the oil formation volume factor was about two and
a third (Hsieh 2010). This does not represent the actual observed change between
reservoir volume and surface spill volume since dissolved gas release, other weathering processes, and incorporation of surrounding seawater are not included.
Like most fluids, the density of the oil increases as the temperature decreases.
The increase parameter is a nonlinear function of temperature and density (ASTM
2007) but can be approximated as linear over conditions outside the reservoir.
Another important bulk property is viscosity. Unfortunately, the term relates to
two different properties with different dimensional units. Kinematic viscosity has
dimensions of area/time with its SI unit being the stoke. Dynamic viscosity, sometimes called absolute viscosity, is kinematic density multiplied by the oil density. Its
SI unit is the poise. The oil industry traditionally used neither but instead would use
the time for an oil sample to flow through a certain type of measuring viscometer
(e.g., Saybolt universal second). Fortunately, this is less common today, and most
large oil property libraries store viscosity in one hundredth of poise or centipoise.
Oil viscosity is highly sensitive to temperature change. Past practice in older
surface spill models was to utilize the Eyring’s equation (1935) to calculate (kinematic) viscosity
v v
k T T
v
0
0
1
1
=
−
ref
ref
exp
(2.2)
at the environmental temperature T 0 (K) by extrapolating from some measured laboratory viscosity v ref and temperature T ref . Choice of k v varied depending on the model,
but a typical value, expressed in Kelvin, would be 5000 K (Bobra and Callaghan
1990). Industry itself needed more accurate estimates over greater environmental
extremes, so more complex methods have been developed (e.g., Orbey and Sandler
1993; Abu-Eishah 1999; Hemmati-Sarapardeh et al. 2013) but are not necessarily
imported into current spill modeling.
While mass loss through evaporation (for surface oil) or dissolution will also increase
viscosity, the most significant cause of viscosity increase for susceptible oils is water-inoil emulsification where the emulsified oil viscosity can increase by more than an order
2 The Importance of Understanding Fundamental Physics and Chemistry of Deep Oil…