March 4, 2016

A Water Resource Risk Assessment for a Highway


A Water Resource Risk Assessment for a Highway: Construction and Operational Phases and Accidental Spills

Lenmour Bell1, BSc; Kris Freeman1, BSc; Christopher Burgess, MSc. PE1, Carlton Campbell, MPhil2

Hydrogeological recharge and contaminant fate model were used to investigate the risks of new 47km highway that traverses important aquifer ground water resources, including wells, sinkholes, caves and rivers. The vulnerability of the population served and critical social facilities were assessed. Mitigating control measures were specified and include wet ponds, vegetated swales and sinkhole treatment which can be used during the operational phase.

1. Introduction

A water resource risk assessment essentially is the evaluation of the hazards to a water resource in order to minimize or totally eliminate the risks involved. It seeks to identify groups vulnerable to the effects of water resource contamination.  An assessment was done on the Highway 2000 North-South Link project to assess the potential impact of the highway on the water resource. This involved identifying the water resources surrounding the highway alignment and determining the hazards and risks that it poses to these resources.

The highway project involved the construction of a 20 m right of way with a further 100m road reservation. The highway is expected to extend a total of 44.9 Km (Section 1 being 28.7 km and Section 3 being 16 km) amassing 45 crossings of streams, rivers and gullies.

The hazards associated with this water resource assessment can be divided into two (2) phases: the construction phase and the operational phase of the highway. Each phase has its characteristic hazards and associated level of risk.

Table 1 summary table of the hazards identified

Construction Phase Operational Phase Accidental Spill Scenario
Site clearing Vehicular operation Gasoline
Drainage pattern modification Atmospheric deposition Sodium hydroxide
Petroleum chemicals used in paving Surrounding land use Sulfuric acid


2. Method

The approach taken in this assessment involved both qualitative and quantitative methods of data collection and analysis. The main aspects of the water resource risk assessment include the following:

  1. Determination of the importance of water resources (ground and surface water) in the study area
  2. Assessment of the vulnerability of these water resources (quantity and quality).
  3. Contaminant transport modelling from both operational phases under normal circumstances as well as accidental spills.
  4. Assessment of the flooding vulnerability of lands, existing settlements and other infrastructure, adjacent  to the proposed road alignment as determined by topography, geology, soil and typical highway drainage system designs.
  5. Evaluation and recommendation of mitigation/control measures which include engineering interventions and appropriate management practices.

Hazardous threats posed by the highway were identified. These included:

Contaminant transport modeling was undertaken before the risk assessment to determine the consequences of such operational and spill scenarios. Groundwater Modeling Software (GMS) is a comprehensive graphical user environment for performing groundwater simulations. GMS was designed as a comprehensive modeling environment where several types of models are supported and facilities are provided to share information between different models and data types. Tools are provided for site characterization, model conceptualization, mesh and grid generation, geostatistics, and post-processing via numerical models. Numerical models such as MODFLOW and MD335 Numerical models are programs (packages) that are separate from but incorporated within GMS that were used to simulate the groundwater flow of water in the study area.

The modeling process required the input of parameters such as well lithology, pumping rates at the wells and the recharge rate of the area. The well lithologies gave the various soil layers and their depths and also the water table heights. With this data it was possible to create a representational grid of the existing ground. Proper calibration is crucial and involved systematic alteration of parameters and the model was repeatedly run until the computed solution matches field-observed values (water table elevations obtained from well logs) within an acceptable level of accuracy. The purpose of this modeling is to better understand the groundwater network. With the results of the modeling, the potential for contaminant transport from the highway over the aquifer to wells that utilize the aquifers were examined.

The final part of the risk assessment is the development of a risk matrix. The matrix formulated was used to prioritize the hazards identified. The matrix involved:

  1. Defining the population to be affected by a particular hazard
  2. Define hazards and likelihood of occurrence,
  3. Define the level of certainty/ uncertainty of its occurrence,
  4. Calculating the risk associated with each hazard (as a population and as a cost) and
  5. Rank the risk in order of significance

The prioritization of the hazard is helpful in ensuring that significant hazards are dealt with accordingly. Appropriate control and mitigation measures were then identified for implementation.

3. Data

Wells in close proximity to the alignment are at risk of being contaminated due the discharge and seepage of contaminated highway runoff into the ground and aquifer beneath. A well database consisting of all wells including production, exploration and observation wells was provided by the Water Resource Authority (WRA). The highway traverses two different hydrologic basins: Section 1 traverses the Rio Cobre basin and Section 3 traverses the Dry Harbour Mountain Basin.

Table 2 Summary table of wells located within each basin and number of wells at risk

Hydrological basin No. Wells in basin Wells at risk
Section 1 Rio Cobre 518 8
Section 3 Dry Harbour Mountains 100 3


Recharge describes the hydrological process whereby water moves from the surface downward to the ground. Flow into the hydrological system is due to infiltration from precipitation. Flow out of the system is due to discharging wells represented by a constant head boundary. The recharge values were obtained from the Water Resource Authority (WRA) Master Plan as summarized below in Table 3.

Table 3 Summary of properties of the hydrological basins

Hydrologic Basin Area (km2) Rainfall (106m3) Evapo-transpiration (106m3) Surface water runoff (106m3) Groundwater discharge (106m3/yr)
Río Cobre 1,283 2,009 1,450 177 472
Dry Harbour Mountains 1,362 2,450 1,302 457 691


The analysis of the DEM and depressions mapped from the 1:12,500 map of Jamaica revealed an estimated number of sinkholes to be one (1) in Section 1 and eight (8) in Section 3. It must be noted that the science of this estimation is only limited to sinkholes exposed at the surface: therefore other sinkholes may be present beneath the surface.

Table 4 Summary table of depressions and sinkholes

Number of depressions from GIS DEM within 1km of alignment GIS sinkholes overlapping mapped depressions Actual sinkholes mapped
Section 1 250 10 1
Section 3 120 50 8


The rainfall experienced in areas where the alignment traverse was determined by analyzing a map of the spatial distribution of rainfall over the island. Based on the spatial distribution of mean rainfall shown in Figure 1 , Section 1 of the highway experiences a yearly mean rainfall of 1,250mm to 1,500mm while Section 3 of the alignment experiences 2,000 mm to 2,500 mm of rainfall per year.

Figure 1 Map of spatial distribution of mean yearly rainfall (1992 – 2012)

4. Typical highway contaminants

Typical pollutant loading on highways is shown below in Table 5. The loadings are given in units of kilogram per hectare per event (kg/ha/event) and represent the load of pollutant that would accumulate on the road surface and be washed off with each rain event.

Table 5 Contaminant Loading (Source: Barrett et al, 1994)


5. Results and Findings

Contaminant concentrations in the operational phase were determined for an event and compared with the water standards discussed earlier. The results indicated that majority (
94%) of the contaminants are likely to exceed the standards in the first foul flush.

Water infrastructure vulnerability is defined as the surface and groundwater resources that are exposed to the hazards. The importance of the water resources in the two sections cannot be understated. The total estimated vulnerable population is approximately 525,000 persons (2001 census). Each well is estimated to provide for approximately 30,000 persons. This estimation was done by considering a per person consumption rate of 50 IGD.

Figure 5 Graph showing well location relative to road reservation and the estimated population served

Ground water modelling using GMS for a transient period of three hundred and sixty-five (365) days, produced pathlines of water movement to well. The lengths of each path line indicate the transmissivity of the underlying soils which has definitive properties such as porosity and conductivity.

Figure 6 GMS output plot of well pathlines for section 1 and section 2 respectively

In Section 1, the zone of influence of these wells was observed to extend to the north. Due to the steep mountainous regions north of the wells, the groundwater that supplies the wells flow in a south-easterly direction towards the pumping location of the wells. As for Section 3, the zone of influence was observed to extend south. The steep mountainous regions south of the wells also playing a major role in influencing the groundwater. A closer look at Figure 6 reveals an interaction between the well pathlines and the highway alignment. This indicates that any release of contaminant on the highway would be able to enter the path line and ultimately end up in the well. This is more evident in section 1 where there are more path lines intersecting the alignment.

Contaminant transport was then modeled using a particle tracking model. The model was executed for a transient period of three hundred and sixty-five days (365) days and the movements and concentrations of the contaminant observed. Oil and grease was modeled in the operational phase and benzene in the accidental spill scenario because these contaminants had the highest required dilution. The locations chosen to model were low points along the highway profile elevation as well as areas with wells nearby. The results of the model were plotted and are shown in Table 7 below.

Table 7 Summary table of output plots of contaminant transport.

The results for oil and grease showed that in Section 1, At the end of the simulation (Day 365), the concentration observed at the nearest affected well was increased to a value of over 0.76 mg/L (EPA safe drinking water standard is 0.01 mg/l). The least affected well within the band of runoff had a concentration of 0.03 mg/L. A total of 4 wells would be affected by this contamination. In section 3, however, the contaminant would have spread further away from the source but at lower concentrations. At the end of the simulation, the concentration observed at the nearest well had value of 0.01006 mg/L.

The results for benzene were similar to that of oil and grease with section 3 having a wider spread than section 1. The highest concentration of contaminant in section 1 was 0.01 mg/l while in section 3 it was 0.2 mg/l after 1 month. These concentrations were further reduced to 0.45mg/l and 1.4 mg/l respectively.

The result produced used a conservative or non-decaying constituent and believed to be appropriate in order to develop an understanding for the potential issues that operational and accidental pollutants can have on the water resource assets. However the approach was  somewhat limited by the number of core hole lithology, well pumping rates and regional pumping studies data that could be accessed at the time.

The risk assessment matrix revealed that during the construction phase, the highest rank risk in terms of number of persons affected would be that of silted runoff reaching to sinkholes. However, the highest ranked risk in terms of a monetary figure would be that of spills of petroleum and other construction chemicals as it would prove more expensive to replace a contaminated well. During the operational phase the highest ranked risk in terms of population that would be affected, is that of the contamination of highway runoff with oil and grease. Although it occurs in small quantities on the highway, it requires a higher dilution to reach acceptable standards as well as it has the potential to impact the entire aquifer. These ranks help to prioritize the implementation of control measures.

6. Conclusions

The water resource risk assessment revealed potential hazards and their risk to the vulnerable society. The risk assessment prioritized these hazards and helped in understanding the level of effects to be had if left uncontrolled. Various control measures were determined that would prove advantageous to the water resources. The control measures include the both engineering measures of control as well as mitigation measure.

The engineering solutions included the use of wet ponds, vegetated swales and sinkhole treatment. Wet ponds are storm water facilities constructed to provide storage of contaminated runoff. Wet ponds consist of a permanent pool of standing water that promotes for gravitational settling, biological uptake and microbial activity. These wet ponds were sized to hold and treat the first foul flush and placed at low points along the highway alignment as these areas are expected to be points of storm water discharge.


Leave a Reply

Your email address will not be published. Required fields are marked *