Appendix 1: Detailed Description of the Exposure Scenario for Release of Quinoline in Water

Appendix 1

Figure A1.1 Generic abandoned gasworks site. (A) “Aerial” view showing the extent of coal tar contamination. (B) Geological cross-section showing the zone of contact of the aqueous plume with the river bottom. (C) Scenario used to define the spatial pattern of contamination in the river. A mixing box is used to model the contamination of river water by quinoline following initial contact with contaminated groundwater. Panels (A) and (B) are adapted from a case study in Ontario reported by Raven and Beck (1992); a review of information from over 100 abandoned sites in the United States has also been used to define the generic site (GRI 1990). About their case study, Raven and Beck (1992) wrote that “because the zone of non-aqueous phase contamination extends to the river, discharges of PAH (this includes azaarenes)…contaminated groundwater to the river will occur at this site for several decades.” Panel (C) is not to scale.

ChemSim Simulations

Substance Evaluated

Substance: Quinoline

CAS RN: 91-22-5

Effluent release type: Plume of groundwater containing quinoline entering a river from bottom sediments

Release quantities: 0.1243, 0.1952, 0.6259 and 0.9777 kg of quinoline per day (depending on quinoline content of coal tar considered; see details below)

No effect threshold: 3.4 µg/L

Model River

Raven and Beck (1992) did not provide any characteristics of the affected stream in their case study. However, in their generic abandoned gasworks site, GRI (1990) defined a river of a width of 11 m adjacent to the generic site. Kettle Creek, in southern Ontario, was a specific example of a river of similar size having been contaminated by manufactured gas operations (OMEE 1997).

River River category – mean flow HYDAT station Latitude/ longitude Data collection period Locality
Kettle Creek Small 02CG002 42.77°N (latitude) 81.21°W (longitude) 1980–2000 St. Thomas (ON)

Channel geometry and hydraulic parameters at this station are as follows: channel width: 14.3 m; mean flow depth: 0.29 m; mean flow velocity: 0.30 m/s.

Loadings

Releases of quinoline to the model river are based on a case study in Ontario in which a large area of non-aqueous-phase pool of coal tar extended towards a river next to a gasworks site (Raven and Beck 1992). The parameters that follow were used to derive quinoline loadings:

The exposure scenario considered the formation of a contaminated groundwater plume containing quinoline in contact with a pure coal tar phase in the soil (Figure A1.1 above). Raven and Beck (1992) qualified this situation as chronic; consequently, we assumed that steady state was reached with all sorption sites fully saturated with respect to quinoline. The following equations were used:

  1. Dissolution of quinoline in groundwater according to Raoult’s Law: Max Ci = xi× Cswi, where xi = weight fraction of the component in the tar, 0.0011 and 0.005 65 above, and Cswi = solubility of the component in water. Ci is expressed as g/m3.
  2. Contaminant transfer at the source, i.e., in contact with the coal tar plume: F = qCi, where q = vn; v is groundwater velocity, 0.09 m/day, and n is soil porosity. The equation has been obtained from King and Barker (1999). F is expressed in units of g/m2 per day.
  3. Contaminant transfer at the sediment–water interface: the equation in 2), F = qCi, was used. The average distance from the coal tar plume to the sediment–water interface was 12 m. Therefore, lateral dispersion was assumed to be negligible. Aerobic biodegradation was assumed to affect 25 m on each side of the 183 m wide plume; as a result, the width of the non-aqueous plume was reduced to 133 m. The centre of the plume was assumed to be under anaerobic conditions not conducive to biodegradation, as suggested by the field experiment of Fowler et al. (1994) with coal tar creosote.
  4. A “mixing box” was superimposed on the contaminated groundwater plume on the river bottom. The volume of the box was the area, adjusted for the river flow above, multiplied by a water column height of 0.05 m. This approach took into account 1) the requirement that the plume be modelled like a diffuser-type source rather than like an end-of-pipe release by ChemSim, 2) the fact that following diffusion through the sediment–water interface, quinoline would remain near the river bottom because its density is higher than that of water, and 3) the fact that an uncontaminated volume of water would see its quinoline content increasing steadily while passing over the contaminated river bottom. The mixing box was divided in subvolumes of 1 m ´ 1 m ´ 0.05 m in order to derive a cumulative mass of quinoline at the end of the box (i.e., entry value for ChemSim in a diffuser-type pattern) and an average concentration of quinoline for the entire box. The ChemSim model calculated the concentration of quinoline assuming instantaneous dilution, which is a less conservative scenario than the one based on the formation of a plume developing from the diffuser-type source. Aerobic biodegradation was accounted for in these simulations. Four estimates of daily input of quinoline into the mixing box (kg/day) were calculated:
50% flow 10% flow
x1 = 0.0011 0.1952 0.1243
x2 = 0.005 65 0.9777 0.6259

Output Summary

Table A1.1: Summary of the ChemSim Output Data
Model river: Kettle Creek,
St. Thomas (ON)
10th-percentile flow 50th-percentile flow
x1 = 0.0011 x2 = 0.005 65 x1 = 0.0011 x2 = 0.005 65
Stream flow (m3/s) 0.14 0.14 1.07 1.07
Quinoline input into mixing zone (kg/day) 0.1243 0.6259 0.1952 0.9777
Dissolved quinoline concentration in surface water in the mixing zone, assuming instantaneous dilution (µg/L) 10.3 51.7 2.11 10.6

ChemSim is a geographic information system–based aquatic exposure estimation model designed to estimate the dispersion and transport of substances released to watercourses. ChemSim combines estimated release quantities with information regarding the receiving watercourses to estimate aquatic exposure values. The estimated exposure values are characterized in the following three ways:

  1. Concentrations of substances within the mixing zone (i.e., plume) can be predicted.
  2. Percentage of the river width affected by the plume can be estimated.
  3. Area of the watercourse with concentrations greater than a specified threshold can be estimated.

ChemSim was developed by the Canadian Hydraulics Centre of the National Research Council Canada and Environment Canada’s National Water Research Institute.

Page details

Date modified: