Nonaqueous Fluids in the Vadose Zone

1
Nonaqueous Fluids in the Vadose Zone

• A brief overview of a messy topic

2
Nonaqueous Fluids in the Vadose Zone

• Much vadose study aimed at contaminant transport
• One set of contaminates requires special
  treatment
• those that are not miscible in water.
• referred to as Non-Aqueous Phase Liquids NAPLs,
  
• low solubility in water.
• non-polar compounds which remain as separate
  liquid phase (as opposed to alcohol or latex).
• Subdivided into those with density
• lower than that of water (LNAPLs - Light e.g.,
  gasoline)
• denser than water (DNAPL - Dense, e.g., TCE,
  carbon tetrachloride).

3
Numerous sources - LNAPLs

• Most ubiquitous
• leaking underground storage tanks (LUSTs)
• Gas stations
• 10 of single walled steel tanks leaked,
• plumbing leaks in approximately 30 of these
  installations
• lesson dont assume that the plume will be under
  the tank since most arise from delivery system
  failure (Selker, 1991).
• Note Most commercial single walled USTs have
  been removed in the U.S. due to tightened
  regulation.

4
Sources - LNAPLs cont.

• Major source of LNAPLs household heating oil
  tanks.
• Long overlooked, there are a vast number of
  leaking buried oil tanks, (same proportions as
  old gas station tanks)
• Household leaks rarely noticed until catastrophic
  failure, since there are no records of
  consumption.
• The lower volatility of heating oil also limits
  the observation of leaks through vapor transport
  into basements etc.

5
Sources - DNAPLs

• DNAPLs in the environment typically arise from
  disposal of cleaning compounds.
• Whereas LNAPLs are most commonly observed at
  points of delivery, DNAPLs are found at points of
  delivery, use, and disposal.
• Dry wells and other ad hoc disposal sites
  represent a major portion of plume generators,
  often near the point of use, or at waste disposal
  sites.
• Spills are typically of smaller volume than
  LNAPLs, but more serious due to higher toxicity
  and bulk penetration of aquifers

6
A typical scene
7
The Components of a Plume
8
The Anatomy of a NAPL Spill

• Prediction of NAPL movement complicated by
  physical and chemical processes making
  quantitative prediction generally impossible for
  field spills (Osborne and Sykes, 1986 Cary et
  al., 1989b Essaid et al., 1993).
• Most productive to understand the qualitative
  characteristics movement, rather than spend
  inordinate energy on quantitative prediction of
  NAPL disposition.
• A key point residual saturation can account for
  a large fraction of a spill.

9
Influence of Water Table
10
Permeability
11
Residual NAPL

• NAPLs tend to form small droplets (a.k.a.
  ganglia) in the unsaturated zone
• On the order of 5 of the volume of the region
  which experienced NAPL transport will remain NAPL
  filled with residual product (Cary et al.,
  1989c)
• This important for planning in soil clean up, as
  well as understanding how much of the product may
  have reached the upper aquifer.

12
Example of residual

• A spill of 10,000 l of product 10 m above an
  unconfined aquifer. Assuming that the NAPL
  wetted area of 4 m by 4 m and a residual
  saturation of 5, how much of this original spill
  makes it to the water table in liquid form?
• Solution
• The residual volume in the vadose zone is
• 10 m x 4 m x 4 m x 5 8 m3
• 8,000 l
• therefore about 2,000 liters (20) makes it to
  the water table.
• Obviously our uncertainty exceeds /- 20, so we
  really have little idea of how much made it to
  the water table, but should assume that a
  significant amount did.

13
Geologic Effects

• Geologic configuration key to disposition of
  NAPLs
• LNAPLs the vadose zone is of primary importance,
  since the bulk liquid does not penetrate the
  saturated zone,
• DNAPLs the structure in both saturated and
  unsaturated regions will have a major impact on
  disposition.
• Main issue layers between media of different
  texture. In particular, horizontal bedding
  features will cause the plume to spread laterally
  with a dominant down-dip movement (Schroth et
  al., 1997).

14
Geologic Effects
15
Real Data(Kueper et al., 1993)
16
Rate of introduction highly influential

• Rapid spills
• require broader areas to carry the flow
• larger residual saturation in the unsaturated
  zone
• less free product on aquifers
• less susceptible to extreme lateral flow due to
  textural interfaces.
• Slow leaks
• more susceptible to lateral diversion along
  textural interfaces
• likely follow more isolated paths of flow
• Slow leaks tend to contaminate a larger area,
  while still delivering a greater fraction of the
  product to the aquifer

17
Rate of spill effects
18
Real Data (Kueper et al., 1992)

• The upper plot is from
• an instantaneous
• release, while the lower
• plot resulted from a
• slow injection, which
• penetrated further, and
• spread more widely

19
LNAPLs vs DNAPLs

• In the vadose zone DNAPLs and LNAPLs behave quite
  similarly if saturation not encountered.
• Logical since the only distinction we have made
  between these is their relative density in
  comparison to water.
• there are no buoyancy effects in vadose zone
• the physics of flow is essentially the same
• Once saturated regions encountered, migration
  differs dramatically for LNAPLs and DNAPLs.
• LNAPLs travel in direction of the slope of the
  water table
• DNAPLs travel in direction of slope of the lower
  boundary
• DNAPLs move through aquifers in web like networks
  of pores (e.g., Held and Illangasekare, 1995).
• this reduces residual saturation, thus increasing
  the free product available to spread through the
  aquifer.

20
LNAPLs vs DNAPLs
21
DNAPL Migration
22
DNAPL Migration
23
Observing LNAPLs in Wells

• Often the first indication of NAPL contamination
  is the observation of the product in a well
• The extent of a plume at a site is often then
  delineated by installing additional wells on the
  site
• The extent of contamination is then delineated by
  obtaining core samples and observing the depth of
  "free product" in the wells
• BE CAREFUL The depth observed in wells is not
  the free product depth on the aquifer

24
Geometry of LNAPLs in wells

• Typical observation well at an LNAPL spill site
  where Hoil is the True depth of free product,
  Hcap is the thickness of the capillary fringe,
  Happ is the apparent depth of free product, and
  Hd the depression of the water surface in the well

25
Calculating some depths

• At the oil-water interface in the well, the total
  head is
• the total head at all points in the aquifer is
  constant (assuming that we are not pumping from
  the well), so head at the interface is also given
  by
• Equating these we obtain

26
Finishing the algebra

• From the set-up geometry
• solving for Hd
• We may rewrite this using the geometric result as
• Solving for Hoil
• NOTE
• denominator
• small!

27
Example

• For typical NAPLs goil/gw) is about 0.8.
  Taking Hcap to be 50 cm (typical for a silt loam
  texture), and assuming the true depth of free
  product to be 2 cm, we can use 2.162 to
  calculate the apparent depth of NAPL in the
  well
• almost 3 m of free product in the well!
• Very sensitive to
• the height of the capillary fringe
• the density contrast of the liquids
• density contrast easy, but the height of the
  effective capillary fringe is difficult to
  measure.

28
Data from experiments

• Observed Actual
• in well free product

29
DNAPLs and wells...

• In the case of DNAPLs, wells present a more
  serious threat.
• If a well screen crosses an aquitard, the well
  itself can become a pathway for transport, with a
  DNAPL draining off the aquitard, into the well,
  and out the well into the lower aquifer.
• For LNAPLs, by creating a cone of depression
  about a well you may facilitate removal of the
  contaminant which will then flow to the well

30
DNAPLs in Wells
31
Movement and Retention

• 1. Initial emplacement
• 2. Soluble losses
• 3. Aging

32
Initial Emplacement

• We have already discussed the over-riding issues.
  A few more remarks
• Movement strongly affected by surface tension
• Surface tension is a function of TIME!!
• changes rapidly in first hours as interfaces come
  to local equilibrium with fluids (on the order of
  30 change)
• changes slowly as the fluids age through
  partitioning losses
• changes slowly as local microbes put out
  surfactants
• Movement typically unstable. No codes handle
  this.
• Any predictions must be field validated

33
Textural Interfaces Multiphase flow

• Lets look at three oil spill cases
• no water flowing
• little water flowing
• lots of water flowing

34
Soluble losses and aging

• Many NAPLs are moderately soluble in water
• Since there is much more water than NAPL, this
  leads to significant losses (plume)
• Many NAPLs are mixtures of hydrocarbons etc.
  (e.g., gasoline has 10s of major components)
• Each of the constituents will partition into the
  water and gas phases according to its own
  solubility
• As the NAPL sits, it changes it makeup becoming
  less soluble/volatile (aging)

35
Partitioning of Common NAPLs
36
Skimming Free Product
37
Summary on NAPLs

• Understanding the physics and chemistry of NAPL
  movement is helpful
• Dont expect to accurately predict disposition
• This has only been a brief overview. Lots of
  very good work on these issues