Tải bản đầy đủ - 0 (trang)
VII. Transport of Viruses in Porous Media

VII. Transport of Viruses in Porous Media

Tải bản đầy đủ - 0trang

SUBSURFACE VIRUS FATE AND TRANSPORT



71



viruses can move considerable distances through the vadose as well as through the

saturated zones (Rossi et al., 1994; Sinton et al., 1997; Yates and Yates, 1988).

Such findings have raised concerns about groundwater contamination. Regulations

regarding drinking-water well disinfections (USEPA, 2000) have led to increased

efforts in past years to elucidate the mechanisms of virus transport through soils

and aquifers. Different aspects of virus transport through soil and groundwater

have been reviewed (Gerba and Goyal, 1981; Keswick and Gerba, 1980; Melnick

and Gerba, 1980; Schijven and Hassanizadeh, 2000; Yates and Yates, 1988).



A. MECHANISMS

The movement of viruses, as well as other microorganisms, through the subsurface is dominated by advection and dispersion and has been modeled using various

forms of the advection–dispersion equation (ADE).

Virus transport through porous media is influenced by adsorption, desorption,

and inactivation processes. These processes are in turn affected by the various

factors listed in Table I, many of which have been the focus of virus transport studies

and are discussed in the following. Column studies and experimental conditions

are listed in Table VI.

1. Soil Properties

Virus retention and transport have been found to be affected by soil properties. In

a set of saturated column experiments, the behavior of three viruses (poliovirus 1,

reovirus 3, and φX174) was studied in several different soils (Funderburg et al.,

1981). A high poliovirus concentration measured in the column percolates correlated most favorably with low soil cation-exchange capacity and high organic

carbon and clay content, whereas a high percolate concentration of φX174 was

related to low soil organic carbon content and residence time of liquid within a

column in combination with either high soil pH or percentage of clay. As with

poliovirus, the detection of reovirus in the soil column percolates was negatively

correlated with soil cation-exchange capacity. These experiments were conducted

with field soils in laboratory columns, which were packed in such a manner that

they simulated the vertical profile and bulk density of the soil as found in the field.

However, the procedures used for saturating the columns and applying input solution were not well controlled, so that complete saturation and constant steady-state

flow conditions were not guaranteed. In addition, because of the complex nature of

the soils used, such experiments do not allow careful evaluation of specific mechanisms. Because of these limitations, results from most of the earlier studies are

difficult to interpret and quantify (Drewry and Eliassen, 1968; Gerba and Lance,

1978; Landry et al., 1979).



Table VI

Transport of Viruses in Columns

Column



Porous material



Virus



Condition



Background electrolyte



Diameter

(cm)



Length

(cm)



MS-2



Sat



Groundwater (pH 8.1)



5



105



Fractured tuff



f2



Sat



Groundwater (pH 8.1)



6.5



25



Silica beads



MS-2, PRD-1



Sat



0.9



Silica beads



MS-2, poliovirus 1



Sat



Ottawa sand



MS-2, φX174



Sat



Accusand

(water-washed)

Accusand

(oxide-removed)

Aquifer material



MS-2, φX174



Sat/unsat



MS-2, φX174



Sat/unsat



MS-2, PRD-1, Qβ,

φX174, PM2



Sat



Phosphate-buffered NaCl

(pH 5)

Phosphate-buffered NaCl

(pH 5.7–8.2, ISa 0.5 M)

Phosphate-buffered saline

(pH 7.5, IS 0.16 M,

0.002 M)

Artificial groundwater

(pH 7.5, IS 0.002 M)

Phosphate-buffered saline

(pH 7.5, IS 0.16 M)

Phosphate-buffered saline

(pH 7.5, IS 0.16 M)

Groundwater (pH 7.1)



References

Bales et al. (1989)



15



48.6

22.8, 28.2

187

225

13.3



0.9



15



19.6–28.1



Bales et al. (1993)



7.6



10.5



2.8–3.4



Chu et al. (2000)



7.6



10.5



3.4, 4.8, 16.1



Chu et al. (2001)



7.6



10.5



3.3, 23.3, 13.9



Chu et al. (2001)



5



76



ns



Dowd et al. (1998)



72



Sand



Pore water

velocity (cm h−1)



Bales et al. (1991)



T1, T2



Sat



nsb



2.86



45–50



0.08–0.33



Loamy sand



Poliovirus 1



Sat



sec



5.6



19.5



ns



Eight soils



φX174



Sat



se (pH 7.2)



10



33, 66, 100



ns



Loamy sand



Poliovirus 1

Reovirus 3

Poliovirus 1



Sat

se



ns



10



250



6.3



Ottawa sand



MS-2, φX174



Sat



9.2



10.5, 20



2.5, 10



Ottawa sand



MS-2



Sat



7.6



10.5



2.6



Jin et al. (2000b)



Ottawa sand



MS-2, φX174



Sat/unsat



7.6



10.5



3.8–18.4



Jin et al. (2000a)



Sand, sandy loam

Sand



Coxsackievirus B3

MS-2, PRD-1



Sat/unsat

Sat



ns

2.7



30, 100

10.6, 14.8



ns

8.6–11.2



Sand



Poliovirus 1

Adenovirus 1



Sat/unsat

Unsat



Phosphate-buffered saline

(pH 7.5, IS 0.16 M)

Phosphate-buffered saline

(pH 7.5, IS 0.16 M)

Phosphate-buffered saline

(pH 7.5, IS 0.16 M)

4.7, 7.5

NaCl (pH 5.7–8.0, IS

0.1–0.2 M)

se

ns



Gerba and Lance

(1978)

Jin et al. (1997)



10

ns



250

100



3.3–10

ns



Jorgensen (1985)

Kinoshita et al.

(1993)

Lance and Gerba

(1984)



73



Five soils



Drewry and

Eliassen (1968)

Duboise et al.

(1976)

Funderburg et al.

(1981)



continues



Table VI—continued

Column



Porous material

Sand, gravel



Virus



Condition



Background electrolyte



Diameter

(cm)



Length

(cm)



Pore water

velocity (cm h−1)



Sat



se (pH 4.4, 7–8.3)



4.3



12.5



∼200



Landry et al.

(1979)



Sat



se (pH 7.0)



2.5



16.5



∼6.5



Lo and Sproul

(1977)

Powelson et al.

(1990)

Powelson et al.

(1991)

Powelson and

Gerba (1994)

Redman et al.

(1997)

Redman et al.

(1999)

Teutsch et al.

(1991)



Silicate



Poliovirus 1, 3

Coxsackievirus B3,

Echovirus 1, 6

Poliovirus 1



Sand

Sand



Poliovirus 1

MS-2



Sat/unsat



Groundwater (pH 8.1)



5.2



105



1–1.3



Flushing Meadows

soil

Sand



MS-2



Unsat



Groundwater (pH 8.1)



5.2



105



1.125



MS-2, PRD-1



Sat/unsat



Se (pH 7.3)



5



100



30–84



Unimin sand



MS-2, Norwalk



Sat



0.01 M NaCl (pH 5 or 7)



1.6



16.9–18.4



90



Sand



SJC3



Sat



ns



19



90



Sand and gravel



T4, MS-2, φX174,

Poliovirus 1,

Rotavirus SA11



ns



NaCl, CaCl2, or MgCl2

(pH 7, IS 0.0003 M)

ns



ns



100



8.3–12.5 or

29.2–33.3



a



IS, ionic strength.

ns, not specified or not available.

c

Sewage effluent.

b



References



SUBSURFACE VIRUS FATE AND TRANSPORT



75



2. Solution Chemistry

The effect of solution chemistry (e.g., pH and ionic strength) has been the focus

of many virus transport experiments. Viruses have variably electrically charged

surfaces, therefore their sorption and transport in porous media are affected by

pH. In general, since porous materials and viruses are both negatively charged

under most natural conditions, increase in pH increases the electrostatic repulsion

between them, thus virus sorption is decreased.

In column studies, higher pH of effluent solution usually results in higher

column-outflow concentration; i.e., fewer viruses are retained in the column.

Bales et al. (1991) reported that a change from pH 5.5 to 8.0 in the eluent solution

produced a large bacteriophage pulse in the outflow, suggesting that increase in

pH was responsible for the large virus desorption pulse during transport. The role

of electrostatic reaction in virus sorption was demonstrated by Bales et al. (1993).

The authors found that the attachment of poliovirus 1 (pHIEP 6.6) to silica sand was

comparable to the attachment of MS-2 (pHIEP 3.9) at pH 5.5, but greater at pH 7

relative to MS-2. A moderate pH dependence was observed on MS-2 detachment

with increasing pH (Kinoshita et al., 1993). In field experiments carried out in a

sandy aquifer, introduction of a pulse of high-pH water at the injection well caused

detachment of viruses (Bales et al., 1995, 1997). The effect of pH on virus attachment/detachment and transport has been confirmed by others (Loveland et al.,

1996; Penrod et al., 1996; Redman et al., 1997; Ryan et al., 1999).

Another important factor affecting virus transport is the ionic strength and composition of the background electrolyte solution. Ionic strength affects the thickness

of the electrical-double-layer surrounding viruses as well as soil particles. An increase in ionic strength shrinks the double layer and provides a closer proximity

between viruses and solid surfaces and therefore enhances virus attachment and

retards transport. Duboise et al. (1976) found that a burst of poliovirus released

from soil columns coincided with a drop in electrical conductivity of the percolate.

The release of MS-2 and poliovirus during transport through silica columns was

enhanced upon changing the eluent ionic strength from 0.5 to 0.005 M (Bales

et al., 1993). The deposition rates of both MS-2 and λ are sensitive to the ionic

strength of the suspending fluid, with more rapid filtration occurring at higher salt

concentrations (Penrod et al., 1996).

The study by Redman et al. (1999) examined the influence of pore water chemistry on the filtration and physiochemical properties of a male-specific filamentous

bacteriophage SJC3. Using a model filtration system consisting of packed columns

of quartz sand, they found that the filtration of this virus was strongly dependent

on the concentration and valence of the dominant cation in the pore fluid. In one

set of experiments, virus retention in the column increased from 0 to almost 100%

when the electrolyte composition of the pore fluid changed from 10 mM NaCl to

10 mM CaCl2.



76



JIN AND FLURY



The composition of both cation and anions also influences virus retention and

transport. Divalent cations were found to be more effective than monovalent cations

in promoting adsorption of viruses to soil materials and wastewater solids (Duboise

et al., 1976; Funderburg et al., 1981; Lance and Gerba, 1984). Anions such as

NO3−, SO42−, and H2PO4− were found to be more effective than Cl−1 in promoting

virus adsorption (Lance and Gerba, 1984). Ryan et al. (1999) showed that adsorbed

phosphate might hinder attachment of PRD-1 and silica colloids to iron oxide

coatings of aquifer materials based on calculations of zeta potentials.

The effects of ionic strength and composition of several buffer solutions on

the inactivation and sorption of two bacteriophages (MS-2 and φX174) were systematically evaluated in a series of experiments conducted using saturated sand

columns (Chu et al., 2000). Changes in ionic strength and composition did not

affect the behavior of φX174, whereas MS-2 was largely removed from a high

ionic strength phosphate buffer solution during transport but moved through the

columns in a low ionic strength phosphate buffer and in an artificial groundwater.

The results indicate that the effects of ionic strength on virus sorption and transport

are strongly virus dependent. Therefore, caution should employed when applying

laboratory results obtained under ideal chemical conditions and using indicator

viruses to the field where pathogens are the concern.

3. Soil Water Content

Soil water content has been found to play a significant role in virus movement

in porous media. Studies have shown that viruses are usually removed more extensively during unsaturated transport than saturated transport (Bitton et al., 1984;

Hurst et al., 1980; Jin et al., 2000a, Jorgensen, 1985; Lance and Gerba, 1984;

Poletika et al., 1995; Powelson et al., 1990; Powelson and Gerba, 1994; Yeager

and O’Brien, 1979). Although the mechanisms by which the water content affects

virus sorption and inactivation during transport are unclear at present, several

possibilities have been proposed in the literature, which are summarized in the

following.

Bitton et al. (1984) and Jorgensen (1985) postulated that the limited virus movement under unsaturated conditions was due to the increased sorption of viruses to

the solid surfaces. Electrostatic and hydrophobic interactions as well as van der

Waals forces are believed to be responsible for virus sorption to the solid–water

interface (Preston and Farrah, 1988). However, Powelson et al. (1990) dismissed

this possibility based on their calculations of the size of water-filled pores that

apparently were much larger than the sizes of viruses. Instead, they concluded

that the strong removal of MS-2 in their unsaturated column experiments was

caused by inactivation, presumably due to the presence of the air–water interface.

Since the sorption of colloids (including hydrophobic and hydrophilic particles of



SUBSURFACE VIRUS FATE AND TRANSPORT



77



clay and polystyrene latex as well as bacteria) to the surface of air bubbles was

directly visualized (Wan and Wilson, 1992, 1994), the presence of AWI has been

suggested by more and more researchers as the dominant mechanism responsible

for the increased removal of colloidal particles, including viruses and bacteria, in

unsaturated systems (Jewett et al., 1999; Jin et al., 2000a; Poletika et al., 1995;

Powelson and Mills, 1996; Schăafer et al., 1998).

The lm-straining theory, introduced by Wan and Tokunaga (1997), proposes

that transport of suspended colloids can be retarded due to physical restrictions

imposed by thin water films in partially saturated porous media. Film straining

becomes effective at a “critical matric potential” and “critical saturation” at which

thick film interconnections between pendular (capillary) rings are broken (Wan

and Tokunaga, 1997). This model predicts that the magnitude of colloid transport

through water films depends on the ratio of particle size (dp) to film thickness (h0)

and on flow velocity. Experiments were conducted using uniform sand and various

sizes of latex particles at different velocities and various water saturation levels,

and results agreed well with model predictions. Experimental results on the effect

of the dp/h0 ratio on the motion of spherical particles in a stable liquid film flowing

down an inclined flat surface were recently reported by Veerapaneni et al. (2000).

They found that (i) at low dp/h0 values, particle velocity increased almost linearly

with increasing particle size; (ii) in the range of dp/h0 = 0.7–1, particle velocity

decreased rapidly with increasing particle size; (iii) in the range of dp/h0 = 1–

1.75, particles ceased to move, and (iv) at dp/h0 > 1.7, particle velocity again

increased with increasing particle size. The authors acknowledged the need to

incorporate the effect of short-range forces (e.g., van der Waals and double-layer

interactions) and the effect of Brownian motion in future work to verify their

reported findings.

Experimental evidence to verify the previously mentioned mechanisms is lacking. Most studies on virus transport through unsaturated porous media have been

conducted under unsteady-state flow conditions (e.g., infiltration experiments),

or in systems with nonuniform water distribution. As such, it is very difficult

to identify and examine the mechanisms responsible for the increased virus removal in unsaturated systems. It should also be noted that previous studies that

demonstrated the importance of the AWI on the retention of colloidal particles in

unsaturated porous media have mostly been conducted under conditions that are

highly favorable for sorption at AWI, such as by using nonreactive solid materials.

In an unsaturated porous medium, it is difficult to isolate the local reactions

of viruses to the solid–water and air–water interfaces. An attempt to differentiate

between the two interfacial reactions has been reported by Chu et al. (2001). These

authors used nonreactive (metal oxides removed) and reactive sands in column experiments under saturated and unsaturated flow conditions. The removal of oxides

from the sand grains rendered the sand inert with respect to virus sorption and



78



JIN AND FLURY



inactivation. The experimental data and a sequential modeling approach showed

that two different viruses (φX174 and MS-2) preferentially sorbed to the solid–

water than to the air–water interface. Virus retardation and removal in unsaturated

columns were mainly controlled by sorption/inactivation at the solid–water interface rather than at the air–water interface. Figure 10 shows this behavior for

φX174, indicating that the lower the water saturation in the columns, the bigger

the difference of virus transport between nonreactive (metal oxides removed) and

reactive sands. Under 100% saturation, φX174 behaved as a conservative tracer

in the nonreactive sand, whereas an irreversible sorption/inactivation occurred in

the reactive sand. Under unsaturated conditions, some sorption to the water–gas

interface was observed, but sorption to the solid–water interface was by far more

dominant.

4. Virus Type

The protein capsids of viruses typically contain ionizable amino acids such as

glutamic acid, aspartic acid, histidine, and tyrosine (Gerba, 1984). Depending on

the pH of the surrounding environment, individual carboxyl and amino groups will

ionize giving the capsid a net electrical charge. Viruses vary in their isoelectric

point. The isoelectric point has been used as one of the most important characteristics in evaluating virus sorption to various solid surfaces. Viruses also differ

in their surface hydrophobicity (Shields and Farrah, 1987). Both electrostatic and

hydrophobic interactions between virus particles and solid surfaces are believed to

control virus adsorption processes (Schijven and Hassanizadeh, 2000), and hence

transport behavior.

Dowd et al. (1998) conducted a study to identify the influence of viral isoelectric

point on viral adsorption onto and transport through a sandy aquifer sediment.

Five different spherical bacteriophages (MS-2, PRD-1, Qβ, φX174, and PM2)

having different isoelectric points (pH 3.9, 4.2, 5.3, 6.6, and 7.3) were used in

laboratory viral transport studies. Conventional batch flowthrough columns, as

well as a continuously recirculating column, in which the outflow is connected to

the inflow, were used. In a 0.78-m batch flowthrough column, the smaller phages

(MS-2, Qβ, and φX174), which had similar diameters, exhibited maximum effluent concentration /initial concentration values that correlated exactly with their

isoelectric points. The amount of viruses sorbed was negatively correlated with

the isoelectric points of the viruses. The data suggest that the isoelectric point of a

virus is the predetermining factor controlling viral adsorption within aquifers for

relatively small viruses (Dowd et al., 1998). However, such a dependence on the

pHIEP was not observed for viruses with a diameter larger than 60 nm (Dowd et al.,

1998).

The filtration behavior of a recombinant Norwalk virus (rNV) and the bacteriophage MS-2, a common surrogate for waterborne viral pathogens, were studied



SUBSURFACE VIRUS FATE AND TRANSPORT



79



Figure 10 φX174 breakthrough curves in water-washed (filled circles) and oxide-removed (open

circles) sand under different water saturations. The solid lines are model simulations considering virus

sorption at solid–water and air–water interfaces. Adapted from Chu et al. (2001). Mechanisms of virus

removal during transport in unsaturated porous media. Water Resour. Res. 37, 253–263. Copyright

[2001] American Geophysical Union. Reproduced/modified by permission of American Geophysical

Union.



80



JIN AND FLURY



in a set of columns packed with pure quartz sand (Redman et al., 1997). The

authors found that in contrast to MS-2, the surface charge of rNV particles and

their filtration through the columns are strongly influenced by pore water pH

over the environmentally important range of pH 5–7 although MS-2 has a lower

isoelectric point. Therefore, MS-2 may not be a good model for the subsurface

filtration of Norwalk virus in natural systems (Redman et al., 1997). Different behavior between MS-2 and φX174 has also been found repeatedly in several studies

(Chu et al., 2000; Jin et al., 2000a; Jin et al., 1997). The two viruses not only

have different isoelectric points (Table II) but also differ significantly in surface

hydrophobicity, with MS-2 being the more hydrophobic and φX174 the more

hydrophilic bacteriophage.

As pointed out by Schijven and Hassanizadeh (2000), virus removal by soil

passage will be virus dependent because of the differences in electrical charge and

hydrophobicity that exist between different types of viruses, and even between

different strains of the same virus type.

5. Size Exclusion

Due to their finite size, macromolecules, viruses, bacteria, and other colloidal

particles are excluded from a certain pore space in a porous medium. Two mechanisms can be considered to be responsible for size exclusion.

1. Particles can be excluded from pores with diameters smaller than the suspended

particle. The particles cannot diffuse into the microporous portion of the medium

and are confined to the larger pores. The microporous region is often a stagnant

phase, so that the size exclusion acts between a mobile and a stationary phase.

This type of exclusion has been denoted as “gel permeation chromatography”

(Casassa, 1971).

2. The center of mass of particles in a single pore is excluded from the immediate

neighborhood of the pore walls due to steric effects. Since the water velocity

is larger in the center of the pore than close to the pore wall, the particle’s

center of mass experiences a larger mean velocity compared to the bulk fluid.

This phenomenon has been referred to as “hydrodynamic chromatography” or

“separation by flow” (DiMarzio and Guttman, 1970; Small, 1974).

In either case, the particles are excluded from regions with no or lower advective

velocities, and it is not really operationally necessary to differentiate between the

two mechanisms (Casassa, 1971). In the absence of chemical reactions, the size

exclusion leads to different mean velocities for particles of different size, with

larger particles usually traveling faster than smaller ones. This phenomenon is

used in preparative and analytical biochemistry to separate polymers of different

size and is commonly known as size-exclusion chromatography (Giddings et al.,



SUBSURFACE VIRUS FATE AND TRANSPORT



81



1968; Yau et al., 1979), without distinguishing between the two different exclusion

mechanisms.

Size exclusion is also known to accelerate the movement of colloids through

soils, sediments, and rocks (de Marsily, 1987; Ginn, 2000; Kretzschmar et al.,

1999). For instance, microorganisms have been found to migrate at velocities

greater than that of the bulk soil water, demonstrating early microbial breakthrough

with respect to a conservative tracer (Crondin, 1987; Harvey, 1988; Harvey et al.,

1989; Mawdsley et al., 1996; McKay et al., 1993). In a microplot field study,

movement of microbial particles such as Cryptosporidium parvum oocysts mainly

occurred through macropores (Mawdsley et al., 1996).

Since size exclusion is used to separate proteins in chromatography columns,

viruses may be affected by this phenomenon as well. In a porous medium, viruses

can therefore migrate faster than a conservative tracer, provided there is no sorption

that would counteract the increased mean travel velocities. The conservative tracer

experiences the entire spectrum of the advective velocities and will therefore have

a smaller mean travel velocity than a size-excluded particle. The magnitude of

the size-exclusion effect depends on the pore size distribution of the medium: the

larger the fraction of the inaccessible pores, the faster the mean transport of the

excluded particle. There is some experimental evidence that size exclusion can

accelerate virus movement in laboratory sand columns (Bales et al., 1989), in

sandy and gravelly soils (Powelson et al., 1993), and in fractured aquifers (McKay

et al., 1993). The acceleration can be substantial: migration velocities for MS-2

and PRD1 in a fractured clay till were up to two orders of magnitude larger than

those for the conservative tracer bromide (McKay et al., 1993).

In addition to the steric size-exclusion phenomenon, viruses may also be excluded from portions of the pore space by electrostatic repulsion. At the natural

pH of most soil water systems, viruses possess a net negative surface charge and

may be repelled by the negatively charged soil particles, a phenomenon known

as anion exclusion. Consequently, viruses will tend to move in the center of soil

pores where the water velocity is greater than near the surface of the soil particle.

6. Colloid-Facilitated Virus Transport

Studies have shown that mobile colloidal particles are generated in soils and

aquifers in situ when changes in solution chemistry occur (Grolimund et al., 1998;

McCarthy and Degueldre, 1993; Ryan and Elimelech, 1996; Ryan and Gschwend,

1994). Suspended colloids can serve as carriers of sorbed contaminants and can

facilitate the transport of strongly sorbing chemicals (Kretzschmar et al., 1999;

McCarthy and Zachara, 1989). Viruses have been found to sorb to mineral surfaces

(Gerba, 1984; Lipson and Stotzky, 1984b), and while associated with suspended

particles they tend to survive longer in the environment (Babich and Stotzky, 1980;



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

VII. Transport of Viruses in Porous Media

Tải bản đầy đủ ngay(0 tr)

×