Category WATER IN ROAD STRUCTURES

Chemical Analysis

7.6.2.1 Introduction

Chemical analysis allows determination of the chemical composition of collected samples and, therefore, to identify specific compounds in the chosen environment. Each chemical compound has one or more analytical methods, from the many dif­ferent methods available, that are more suitable for obtaining an accurate determi­nation of concentration. This section of this chapter presents a brief summary of the analytical methods most used at present for chemical composition identification. It includes coverage of toxicity tests that properly supplement chemical analyses when used to assess the possible impact on living organisms.

7.6.2.2 Selective Ion Measurement

Ion selective electrodes (ISE) are membrane electrodes that respond selectively to specified ions in the presence of other ions. ISE include probes that measure spe­cific ions and gasses in solution. ISE are most commonly used to determine cations and anions. An ISE (with its internal reference electrode, Fig. 7.11) is immersed in an aqueous solution containing the ions to be measured, together with a separate, external reference electrode.

Fig. 7.11 Ion selective electrode – main constituent parts

Ion

Concentration range (mol. l-1)

Ag+/S2-

10-7 ^ 1

Ca2+

5 x 10-7 ^ 1

Cd2+

10-7 ^ 1

Cl-

5 x 10-5 ^ 1

CN-

10-6 ^ 10-2

Cu2+

10-8 ^1

F-

10-6 ^ 1

H+

0 < pH < 14

K+

10-6 ^ 1

NH3

10-6 ^ 1

NO-

6 x 10-6 ^ 1

Pb2+

10-7 ^ 1

Table 7.2 Examples of ion-selective electrodes and measurement ranges

The most commonly used ISE is the pH probe (see Section 7.5.2). Other com­monly used ISEs measure electrical conductivity, metals (see Table 7.2) and gases in solution such as ammonia, carbon dioxide, nitrogen oxide and oxygen.

The principle of the measurement is ion exchange between the ion which is dis­solved in the solution being monitored and the ions behind the membrane Fig. 7.11. The electro-chemical membrane permits the desired ions to cross it, resulting in a charge on the fluid inside the membrane. At the same time the same amount of charge is passed from the reference electrode to the sample solution, thus maintain­ing electrical equilibrium. ISEs are normally available as pen-sized probes that can be lowered into the fluid to be assessed – see Fig. 7.12. An excellent guide to ISEs

and their use is available on-line (Rundle, 2000). Ions commonly analysed using ISEs are listed in Table 7.2.

Chemical analyses are able to give more precise figures but ISEs can be useful to give an approximate value and also indicate a need for more advanced analyses.

Sequential Extraction Methods

Selective extraction can be considered as an “operational speciation” as it corre­sponds to the quantification of elements bound to specific phases of the soil, rather

LEACHING TESTS FOR GRANULAR MATERIALS

pH Domain 4-5

pH 5-6

Material Dictated

Complexation

Low L/S

TCLP

Swiss TVA

DIN 38414 S4

MBLP (Synth)

MBLP

EPtox

NFX-31-210

(California WET

EN 12457-3 (at

test)

L/S =2 & 10)

Availability test

O-norm S2072

Wisconsin SLT

(NEN 7341) California WET

EN 12457

Ontario LEP

Canada EE MCC-3C

Quebec QRsQ

ASTM D 3987

Soil HAc

Soil – NaNO3 Soil – CaCl2

Table 7.1 Examples of leaching (and speciation) tests from around the world (adapted from van der Sloot et al. (1997) and from Hill (2004))

Single Batch Leaching Tests (equilibrium based)

Multiple Batch and Percolation Tests (mostly based on local equilibrium)

Serial Batch (low L/S) Serial Batch L/S>10 UHHamburg NF-X 31-210

WRU WRU

EN 12457-1 ASTM D4793-88

(at L/S = 2) NEN 7349 (NVN 2508)

MEP method 1320 Sweden ENA

MWEP

EN 12457-2 & -4 (at L/S = 10)

Static Methods Speciation Methods

Pacific Northwest Lab. MCC-1 Pacific Northwest Lab. MCC-2 Compacted granular tank leaching test (Rutgers/ECN)

LEACHING TESTS FOR MONOLITHIC MATERIALS

ANSI/ANS 16.1

Tank leaching test NEN 7345 (a static test, non-agitated)

Spray test (impregnated wood)

Swedish MULP

EN 1744-3 (static test, agitated)

The codes in this table refer to various standards or standards originating organisations. Readers who are uncertain of their meaning are referred to the original sources. L/S = Liquid:solid ratio.

than to an exhaustive analysis of the chemical species in the material. (Tessier et al., 1979; Quevauviller et al., 1993). Selective extraction procedures of pollutants from soils can be simple or can be organised according to a sequential or parallel extraction pattern.

Simple extraction procedures are not much used to determine the operational speciation of metals in materials but are used in soil sciences in order to quantify their potential availability for plants.

Sequential and parallel extractions follow the same principle: to submit the ma­terial to a series of reactants in order to identify associations between the different components of the material and the pollutant. Such procedures are more informa­tive than simple extractions as they allow study of the geo-chemical partitioning of pollutants.

Parallel extractions involve different test portions of the same sample subjected to different reactants, while sequential extractions aim to submit the same sam­ple to a well-ordered series of reactants with increasing aggressiveness. Different parallel extraction procedures have been proposed by Seme (1975), Forstner & Patchineelam (1976) and Cazenave (1994) (as cited in Lara-Cazenave, 1994) and also different sequential extraction procedures e. g. Gupta & Chen (1972), Engler et al. (1974), Tessier et al. (1979), Salomons & Fortsner (1980), Meguellati (1982), Welte et al. (1983) and Morrison & Revitt (1987) (as cited in Flores-Rodrigues, 1992).

Associations that are usually studied in most selective extraction protocols are (Colandini, 1997):

• the exchangeable fraction: pollutants are removed from clayey minerals and amorphous materials by simple ion exchange (neutral salts such as MgCl2, BaCl2 or CH3CO2NH4 are used);

• the fraction associated to carbonates: metals (co-)precipitated with natural car­bonates are easily dissolved by a pH decrease (a weak acid as CH3COOH is enough to dissolve calcite and dolomite);

• the fraction associated to metal oxides: metals associated to oxides of Fe, Al and Mn are extracted by means of a reducing agent (as hydroxylamine hydrochlo­ride – NH2OH. HCl);

• the organic fraction: under oxidizing conditions (H2O2 is used under acidic con­ditions) organic compounds are mineralized and metals are released; and

• the residual fraction: includes elements that are naturally present into the matrix of minerals.

In principle the different mineral phases can be quite precisely isolated thanks to the use of a series of extractions. However, the chemical attack on a phase does not always lead to a complete dissolution of pollutants contained in that phase, and for sequential extractions this can result in the dissolution of metals contained in other phases of the sample. Moreover, pollutants released by mineralization of a phase can be reincorporated by remaining phases. Thus, during sequential extraction measuring errors can accrue through the different steps. Despite these drawbacks, there remains a key benefit: that this method requires less material than parallel extraction.

As a conclusion of a European research programme (Ure et al., 1993) a 4-step harmonized sequential extraction method of heavy metals from soils and sediments was proposed by the Bureau Communautaire de Reference: [19]

• extract metals of the oxidizable fraction (using H2O2 8.8 M; CH3COONH4 1M; pH 2); and

• extract the residual fraction (using HF + HCl 15.5 M).

Extraction Through Leaching and Percolation Methods

Leaching can be defined as the process by which soluble constituents are dissolved and filtered through the soil by a percolating fluid, while percolation can be de­scribed as the movement of water downward and radially through subsurface soil layers, continuing downward to groundwater (US EPA, 1997 in ADEME, 1999). This led Tas & Van Leeuwen (1995) to define leachate as water or wastewater that has percolated through a column of soil or solid waste in the environment (in ADEME, 1999). The laboratory terminology describes the leaching test as a technique of leaching of solid products by an appropriate solvent in order to ex­tract its soluble fraction (ADEME, 1999). Leaching tests are a kind of extraction technique.

Extraction may be achieved in a number of ways that may be usefully classified as follows:

• Static tests in which the solid specimen is placed in a container with a fixed volume of fluid (leachant) for a certain period of time during which a static equi­librium is reached between the solid and the solution. Such tests are carried out in few hours. Among them one can distinguish:

о Single batch tests in which the leaching solution is unique. Depending on the test method this may, or may not, involved agitation of some form to quickly reach steady-state conditions. Agitated tests focus on measuring the chemical properties of a material-leachant system rather than the physical, rate-limiting mechanisms. In non-agitated extraction tests the material and leachant are mingled but not agitated: these tests measure the physical, rate- limiting mechanisms.

о Serial batch tests in which a series of single leaching tests is carried out on the same solid specimen. Such an approach, by means of a succession of steady states, is intended to exhaust the total amount of removable pollutant or, at least, to monitor change in leaching with volume of water passing. Leaching tests are generally carried out in few hours. They are simpler to apply but are less realistic than the percolation tests described next.

• Dynamic tests in which, in a column, a continuous supply of fresh leachant is passed through the specimen and withdrawn after contact with the solid fraction. Contrary to static tests, they allow assessment of the release as a function of time. After a while a dynamic equilibrium can be reached generating a continuous release. Such tests can last up to several dozens of days. Among them one can distinguish:

о Up-flow percolation tests in which the column is fed from the bottom. This method implies saturated conditions and avoids preferential flows in the col­umn. It may induce pressure migration. The flow through the column is easier to control than in the down-flow percolation test and this means that it is more often used despite being less representative of usual flow conditions in soils. о Down-flow percolation tests in which the leachant flows under gravity through a partially saturated column. These tests are especially useful in studying bio­chemical activity in the vadose zone (Fig. 7.10).

As many soils are rather impermeable, the test is often accelerated by applying large pressure differentials across the specimen of soil, but this reduces the contact time of the water with the solids, so careful interpretation of results is then necessary to ensure that the laboratory result can be applied to the in-situ conditions with meaning.

In each of these tests, the fluid can be water (often distilled and deionised) or it may seek to be representative of in-situ or “worst-case” groundwater (e. g. a weak acid). For static leaching tests, when the water content of the material is too high (sediments, sludge), interstitial water can be recovered by means of a centrifuge.

The centrifuged pellet is then used to carry out leaching. The centrifuge supernatant may also be subjected to testing.

Chemical mechanisms controlling the release of pollutants are dissolution, sorp­tion and diffusion. Diffusion will be controlled by concentration differences and by the total available contaminant content (which can be far lower than the total con­tent). The pH of the material and its environment (in the laboratory: the solvent) are most important as dissolution of most minerals and sorption processes are pH depen­dent. The oxidation/reduction state of the material and its environment influences the chemical form of the contaminant and its solubility. Complexed forms are gen­erally more soluble than non-complexed ones. The presence of solid and dissolved organic matter or humic substances can enhance the leaching. High ion strength of the solution in the material or in its environment generally increases the leaching of contaminants. Temperature increase leads to higher solubility. Lastly, time of con­tact is an important factor for the release amount (van der Sloot & Dijkstra, 2004).

The form of the material (granular, monolithic or cemented) is an important phys­ical factor influencing the transport of a contaminant from the material to the liquid phase. Indeed, the release behaviour of granular materials is percolation (advection) dominated, while for monolithic materials it is diffusion dominated. For granular materials, the particle size determines the distance between the centre of the par­ticle and the surface area of exchange and also, for a given amount of material, the total exchange surface. The latter factor is also important for monolithic mate­rials, considering the shape of the monolith. For granular (in column) and mono­lithic materials, the porosity and the permeability are important factors on release (van der Sloot & Dijkstra, 2004).

Several parameters can be controlled in leaching test protocols in order to high­light different leaching behaviours:

• the relative amount of solvent in contact with the material (expressed in litres/kilogram of dry material, or sometime in litres/sq. metre for monoliths) or the flow through columns;

• the nature of the solvent (generally de-ionised water);

• the time of contact;

• the pH of the solvent (natural or controlled in order to maintain specific values);

• the granular or the monolithic form of the material;

• the crushing of the material to a certain particle size;

• the porosity and the permeability of compacted granular materials implemented into columns; and

• temperature (which generally is ambient temperature or controlled at 20°C).

Table 7.1 presents some examples of typical leach test methods that can be found in the literature. Also listed, for completeness, are speciation tests that aim to sepa­rate out different leaching species.

Laboratory Measurements

Contaminants may be held both in pore water and on/in the solids fraction of soil samples. Often it is desirable to know how much contaminant could be released from the sample. Simple separation of the pore water (e. g. by a centrifuge method) will not enable us to know how much contaminant might be released from the solids by desorption and leaching. To find this information, extraction tests of some kind need to be performed in which the contaminant is encouraged to move from the solids into the liquid phase by the arrangement of the tests. This is the subject of the first part of this section. Once the liquid phase has been extracted, chemical tests can be performed on the contaminated water – this is described in the second part of this section.

7.6.1 Extraction Methods

7.6.1.1 Introduction

Soils are complex matrices made up of numerous constituents with different and variable physical and chemical properties. Such constituents present variable capac­ities of interactions with pollutants, which drives the partitioning between the liquid and the solid phases. Pollutants of soils can thus be dissolved in the solution, can be adsorbed or make complexes with organic or inorganic constituents, or can be par­tially or totally transformed (bio-geo-chemical dynamic) (ADEME, 1999). All these forms are in relation and can, according to the type of matrix and the properties of the pollutant or external factors, induce an increase or a decrease in the mobility and (bio) availability of pollutants. These different forms can be extracted selectively from the matrix thanks to laboratory extraction methods using appropriate chemical reactants (Tessieretal., 1979; ADEME, 1999).

The environmental performance of a material is rather based on release than on total content of potentially dangerous constituents (van der Sloot & Dijkstra, 2004). Selective extraction procedures allow the assessment of the geo-chemical distribution of pollutants in the solid matrix (Colandini, 1997), and therefore the choice of extraction methods depends on the purpose of the investigation. This section essentially deals with inorganic pollutants.

Organic pollutants are often insoluble in water, though may be miscible by sur­factants (e. g. detergents) or may exist in water as emulsions. Alternatively, the water and the organic chemical may be self-segregating leading to layered “oil” and other fluids, their relative positions dependent on relative density. Interaction of the or­ganic fluid and solid is complex, depending largely on surface chemistry effects which will not be explained here (see e. g., Yong et al. (1992) for further informa­tion). Sorption of organic fluid is largely limited to organic solids.

Redox Potential (in-situ)

The redox potential, i. e. a measure (in volts) of the affinity of a substance for elec­trons compared with hydrogen, may also be determined in the field using electrical, hand-held equipment, this time employing an inert oxidation-reduction electrode.

7.5.2 Electrical Conductivity

Electrical conductivity is typically measured in-situ, being an important, yet sim­ple, indicator of pollution since the ability of water to conduct electricity increases as the proportion of dissolved ions increases. It can be measured directly through the insertion of probes and the resistance (or conductivity recorded) or indirectly through air-coupled “aerials”. However, because there are so many factors that affect electrical conductivity (e. g. presence of metals, saturation), it is normally best to use conductivity techniques as means of locating areas of anomalous response. These can then be investigated by alternative techniques to discover whether pollution is the cause and, if so, its degree and type.

In-situ Measurements

To measure pH in-situ, so-called pH testers (for a rough estimate of the pH value) and pocket (portable) pH meters are used (Fig. 7.9). Periodic calibration of the in­strument is required.

The determination of pH is very fast and reliable when a combined glass elec­trode is used. It enables an automatic measurement over long time intervals with the accuracy of ± 0.01 pH units. The glass electrode can be used even in strongly acidic and alkaline solutions, and also in the presence of oxidizing or reducing substances. It must be constantly immersed in water. With time, all glass electrodes deteriorate due to alkali leaching from the surface layers.

7.5.1 Introduction

When collecting a sample of water, certain principal variables that are prone to more or less rapid change upon sample storage must be measured in-situ. These variables characterize the status of the water at the time of sampling. These variables com­monly include electrical conductivity, pH, temperature, redox potential, and some­times also total hardness, turbidity, salinity and dissolved oxygen. Among the in-situ variables, electrical conductivity determination gives the most important informa­tion about water quality since it gives an indication of the salts dissolved in water.

Ion selective electrodes provide, in principle, a method for users to determine the concentrations of many ions. However, the instruments need careful (and often repeated) calibration to reference concentrations and washing in a buffer solution between successive readings. This makes their routine use on-site somewhat prob­lematic and prevents their sensible use as remote instrumentation. For this reason further details of these instruments are given in Section 7.6.2.

Water and Soil Storage

As physico-chemical and biological reactions occur in the soil and the water, sam­pling periods of short duration are recommended. Some chemical variables should be measured in-situ in a sub-sample (temperature, pH, redox potential, and electrical conductivity) whereas the main water sample is preserved to prevent reduction or loss of target analytes, and transported to the laboratory without delay and kept cool until further treatment. Preservation stabilizes analyte concentrations for a limited period of time. Some samples have a very short holding time (from few hours to some days).

Each analytical method available will have its own requirements for specimen preparation. The most appropriate sampling method specifications for each para­meter can be found in many textbooks (e. g. SMEWW, 1998). They should consider, for each chemical parameter, the bottle type (glass/plastic, dark), preservative (acid­ification, cold), typical sample volume, the need of filtration, and maximum storage time.

Sampling of Soil and Soil Water

Having entered the soil environment near roads, contaminants will either be retained in the soil or transported through the soil. Depending on soil characteristics and other environmental conditions, different contaminants are transported with the soil
water through the soil at varying rate. Mobile compounds (such as chloride) move rapidly whereas many heavy metals and organic contaminants move much slower. Often, contaminant concentrations are much higher in the upper soil layers than further down the soil profile.

Sampling of soil water gives a picture of the rate of transportation of contami­nants down a soil profile whereas sampling of soil gives a picture of the contaminant quantities having accumulated in the various soil layers over a long period of time.

Seepage and soil water (pore-retained water) can be sampled, although with less ease than groundwater below the water table. There, suction lysimeters (also called tension lysimeters) may be used. In principle, the soil water is sucked out of the soil through a lysimeter body that acts as a membrane or filter (Fig. 7.7). These devices include a high air-entry porous tip inside which a partial vacuum may be applied via a flexible pipe connected to an external vacuum pump. By this means water is pulled into the tip and, after collection, is sampled by gravity when possible or by gas displacement. For heavy-metal sampling, lysimeters should be made of Teflon, glass, PET or other material unable to sorb the metals. The flux of contaminants down the soil profile is often of interest. Since tension lysimeters give information on concentrations only, water volumes have to be measured separately or modelled so as to make calculations of pollutant fluxes possible.

Besides the more classical methods of soil water sampling, alternative road sur­face infiltration samplers have been designed in which water seeps from the road surface down and goes through separate layers of pavement and embankment to­wards a circular “funnel” (Sytchev, 1988). The device is installed during pavement construction with the layers of pavement being placed over the top of the sampling inlet. Two layers of siliceous sand of different grain sizes are situated there on fine – mesh screens to prevent entrance of solids to a sampling bottle where the seeping water is collected. The water amount in the sampling bottle is detected by measur­ing resistance (conduction sensor). There are two small metal pieces in the bottle; resistance between them is different when there is air or water. When a sufficient

amount of water has been collected in the sampling bottle, a gas (commonly N2) is injected into the bottle so as to close the valve under the funnel. The water specimen is then forced back to the surface and runs into the sampling bottle (Fig. 7.8a, b).

I = bore hole in road pavement and embankment

2=cement bed 3=sampling bottle (glass)

4 = plug (plastic)

5 = pipe for gas

6 = pipe for water sample 7=sand filter

8=electrical conduction sensor 9=quick-acting coupling (blue for water)

10 = quick-acting coupling (black for gas)

II = connector for conduction sensor

Soil sampling can be performed in any season, except in periods of frost. Sam­pling from consecutive soil depths gives information on the displacement of ac­cumulated pollutants down the soil profile. Natural upper soil profiles are usually richer in organic matter content favouring the retention of several pollutants, namely heavy metals and organic pollutants. This pattern should be analysed in order to observe differences in contaminant content and behaviour across a soil profile. Usu­ally, soil samples are taken out using a steel cylinder of a given volume so as to allow volume-related physical and chemical analyses. Just as for water, soil samples should be transported without delay and kept cool.

To collect samples beneath pavements, a core hole will usually be required in the pavement surface. Drilling conventionally uses a water-cooled core cutter, but these should not be used when abstracting samples for chemical assessments as the water will be likely to change the chemical conditions in the underlying ground by introducing contaminants and/or diluting what was already there.

Sampling of Groundwater

Groundwater sampling can be performed in existing facilities (wells, piezometers, springs, etc.) or in new ones that need to be built. In this latter case, it is more frequent to install piezometers of a small diameter, just enough to be compatible with the monitoring equipment to be subsequently used. Most groundwater sam­pling installations must be located downgradient to the road discharges in terms of the local groundwater flow if they are to look for road-induced contamination. A few upstream locations may also be chosen to allow a reference water condition to be established. Their depth should be at least 2-3 m below the minimum annual groundwater level in order to avoid having a dry piezometer. The necessary number of piezometers depends on the dynamics of groundwater (i. e. varying condition with time and place) (Leitao, 2003). Higher permeability and hydraulic gradient implies monitoring in more locations and more frequently.

To obtain a sample that is representative of the water in the well or piezometer in question, there should be a purging operation until electrical conductivity, pH and temperature of the outflow water have stabilized (Aller et al., 1989). In low permeability soils the purging operation should try to minimize the water displaced during purging, otherwise recharge of the sampling location will take too long to allow realistic sampling. The water can then be sampled directly or using a device similar to Van Dorn’s water bottle, but with a size compatible with the well diameter (Fig. 7.6).

Sampling from Surface Water Bodies

Water samples should be collected from surface water bodies (lakes, rivers, streams, ponds, etc.), taking into account the velocity field in flowing water and any possible stratification in standing water. Once the location and frequency of water sampling are selected there are a set of procedures for sampling water in the natural environ­ment (from surface water bodies) that need to be considered. Figures 7.4 and 7.5 give two examples of bottles for surface water sampling.

1. Ballast

2. Sample container

3. Supporting mesh

4. Rubber stopper

5. Suspension cable

6. Connecting cable

7. Air vent

8. Inlet

9. Vent and inlet caps

The bottle is unsealed at the sampling depth

1. Sample chamber

2. 8 3. Rubber end caps

4. Rubber pull-rod

5. Connecting and locking pin

6. Control mechanism

7. Sample outlet

The sampler is sealed by releasing the two end caps when the sampler is at the correct depth. The rubber pull-rod then contracts, pulling the end caps inwards so that the sampler is sealed at both ends