Earthfx completed an integrated GSFLOW modelling study in support of an application to extract gravel and bedrock from below the water table by James Dick Construction Ltd. (JDCL). Earthfx was tasked with characterizing the current hydrologic and hydrogeologic conditions on site and predicting future impacts of quarrying on Provincially Significant Wetlands (PSW), streamflow, and groundwater levels.
The complex site consisted of three man-made ponds from a previous aggregate pit operation and a southern extraction area, all surrounded by PSWs and flanked by tributaries of Bronte Creek. Extraction was to occur below the water table in such a way to minimize impacts to surrounding hydrologic and hydrogeologic features. Earthfx worked closely with several consultants, including, biologists to develop and optimize an extraction plan while meeting the needs of the client and preserving critical salamander habitat within the PSWs and other sensitive features.
Earthfx employed a local “cutout” of the larger Milton Tier 3 regional GSFLOW model for this study. The local cutout model underwent a re-calibration effort to better match observed water levels at the Reid Road site. An excellent un-coupled steady-state calibration of ±0.5 m was achieved locally at multiple monitoring locations across the proposed site. The integrated transient GSFLOW model was then calibrated to match water levels at Provincial Groundwater Monitoring Network wells and flows at Water Survey of Canada streamflow gauges. The site-specific calibration was then validated with fourteen on-site transient monitors including pond and wetland staff gauges, multi-level groundwater wells, and streamflow gauges. Again, a strong agreement was achieved between simulated and observed data, both in absolute value and in the behaviour of the transient response.
Figure 2 gives an example of the level of calibration achieved at one of the key on-site wetlands (P7A). Matching both absolute water levels and the magnitude and timing of seasonal fluctuations was critical to accurately characterizing current wetland hydroperiod (i.e., number of days a wetland is submerged) and establishing an accurate baseline condition to evaluate impacts of future extraction and closure scenarios against.
A 25-year transient simulation was completed with the calibrated model. Results were post-processed to characterize wetland habitat through hydroperiod analysis and checked against available monitoring data. The simulated water levels relative to the surveyed bottom elevations of the wetlands were calculated daily on a cell-by-cell basis.
It was assumed that inundation occurs when the standing water depth is greater than or equal to 10 cm. Figure 3 shows the average annual hydroperiod in each of the critical wetland features over the 25-year simulation. Results of the analyses indicated that all wetlands exhibited at least some length of hydroperiod; however, not all wetlands were submerged year-round. Many of the features, or parts of the features, exhibited less than 200 days of hydroperiod and tended to dry out in the summer months. The status of the wetlands on July 31st was particularly important for determining the suitability of the wetland for salamander breeding habitat.
Despite no active dewatering occurring during operations, below water table extraction was predicted to adversely affect local groundwater and wetland water levels, thus necessitating mitigation measures. The following measures to minimize the impact of extraction and other water use were considered:
• Limit the rate of below water table extraction to constrain the drawdown within the extraction ponds to 1 m or less. Water may be moved between ponds for supplementation purposes.
• Utilize constructed buffer ponds and dispersion trenches to maintain water levels in key wetlands and to maintain flow in adjacent streams. Water is pumped into buffer ponds or dispersion trenches from the main ponds, while adhering to the specified drawdown limits.
The complex extraction and water management strategy was simulated on a daily basis and compared to the baseline condition. Figure 4 shows simulated water levels in a key wetland (P7A) under baseline and operational conditions. Results showed that extraction lowers the water level in the wetland; however, with supplementation, water levels remained above the hydroperiod inundation threshold on July 31st of each year. These analyses were completed for the seven-key wetlands and showed that there was enough water on site to actively mitigate impacts of extraction. Closure conditions were also simulated and predicted the site to be quite resilient, with long-term pond level and groundwater levels dropping by only 10 to 15 cm. Analysis of post-closure conditions showed that inundated wetland area decreased by an average of 11% on July 31.