Oak Ridges Moraine Aquifer Vulnerability Mapping

Transcription

1 Accompanying Document to the Reference Map for Ontario Regulation 140/02 (Oak Ridges Moraine Conservation Plan) March 2004

2 2 1. Purpose of Report Accompanying Document to the Reference Map for Ontario Regulation 140/02 (Oak Ridges Moraine Conservation Plan) Oak Ridges Moraine Aquifer Vulnerability Mapping This report is the accompanying documentation referred to in the notes of the Reference Map for Ontario Regulation 140/02 (Oak Ridges Moraine Conservation Plan) made under the Oak Ridges Moraine Conservation Act, 2001, dated March, By identifying areas of high aquifer vulnerability, the aquifer vulnerability maps facilitate the implementation of the provisions of Section 29 and Subsection 42(1)(c) of the Oak Ridges Moraine Conservation Plan. 2. Background An understanding of regional groundwater conditions and their inherent vulnerability to contamination is critical to maintaining ecological and sustainable use functions. Aquifer vulnerability mapping is a tool that can be used to protect groundwater resources and their ultimate use. Aquifer vulnerability maps identify areas where contamination of surface water is more or less likely to result in the contamination of groundwater. The aquifer vulnerability maps for the Oak Ridges Moraine (ORM) were prepared with the Ministry of the Environment (MOE) Water Well Log Database. The maps were based on a review of current mapping methods. The mapping method applied was largely computer-based with vulnerabilities calculated on a well-by-well basis and the resulting values interpolated to form a map surface. Regional geological knowledge was incorporated through the use of quaternary geology maps and corrections to the geological descriptions in the well log database. It is important, nevertheless, to remember that the maps do not include a detailed hydrogeological analysis. An underlying theme of this work is that all groundwater is vulnerable to some degree. As such, the terms high aquifer vulnerability and low aquifer vulnerability on the Reference Map are relative characterizations of the state of vulnerability with reference only to one another. In general, aquifer vulnerability maps identify areas where contamination of aquifers is more (or less) likely to occur as a result of surface contamination. It is anticipated that land managers, municipal planners, and facility owners and operators will be able to use the maps showing the areas of high and low vulnerability at specific locations. The rationale for the mapping method is linked to time of travel of water and the contaminants that move in the water (usually in a dissolved state) from the surface to an aquifer. The vulnerability is tied to the arrival of a contaminant at the water table and/or the shallowest aquifer. The mapping method was not geared to assess a specific contaminant, contaminant group or human activity. This method assessed vulnerability with limited consideration of the specific attributes of the hydrogeological system or the behavior of contaminants. The two key attributes considered were the depth to water table and the hydraulic conductivity (K) of geologic material in the unsaturated zone (or above a confined aquifer).

3 3 Fine, un-fractured media retard contaminant migration, whereas fractured or coarse porous media provide faster travel times and less retardation and hence more vulnerability. For example, 20 metres of silt over a confined aquifer would have a low vulnerability, whereas 10 metres of clean, coarse sand or fractured rock would have a high vulnerability to contamination. In the context of the data available from the MOE Water Well Information System (WWIS), aquifer vulnerability can be inferred from the well geology, water table position, vertical gradients and possibly the screen location. The method prescribed here focused on the geology and water levels alone. Screen locations are often considered to be more reflective of a deeper production aquifer, rather then a surficial aquifer with important ecosystem functions. The method adopted was based on calculating a vulnerability index at each well, then mapping the indices using a krigging interpolation technique 1. The index calculation required a sequence of logical statements to evaluate the hydrogeological conditions at each well, followed by statements to sum up the net vulnerability index at each well. 3. Data Sources Data sources for the maps included the following databases and maps: 3.1 Databases The MOE Water Well database served as the source of all well and sub-surface geological information for the map. The database used was from July 2001, and contains approximately 10,000 wells within the ORM boundary. 3.2 Maps Digital Elevation Map (DEM): Ministry of Natural Resources (MNR) based on a 30 m cell size and 10 m contour intervals Ontario Base Maps: MNR, including roads, water courses, municipal boundaries, Conservation Authority (CA) boundaries Quaternary Geology: Geological Survey of Canada (GSC) 3.3 Data Treatment The MOE Water Well Records were the primary source of data used to develop the geological sequence of the ORM and the depths of the aquifers used in the mapping calculations. Characteristics of this data source include: limited confidence in certain aspects of the data (insufficient detail in the geological sequence due to the methodology) and limited well coverage (there are relatively few wells on the eastern section of the ORM) To improve the confidence in the data, in the time available, four main steps were taken: wells with poor positioning quality codes were disregarded 1 Krigging is a geo-statistical method for spatial interpolation whereby existing observations in an area are used to estimate values at un-sampled sites.

4 4 the GSC geo-material conversion was applied the quaternary geology maps were used to define the first geologic unit in each well the geology at each well was simplified into aquifers and aquitards Mapping Calculation Method The selected method was based on the Geology Index, as the Geology Index and Depth Index values are generally similar. The Geology Index is based on summing the thickness of each geologic unit multiplied by its K Number. Like the Depth Index, under confined conditions, the depth to the water table was taken at the depth to the top of the aquifer. Table A offers recommended K Numbers for primary geological materials. For the ORM, this table was modified to reflect the geological conditions of the ORM based on the materials produced by the GSC geo-material conversion. The Depth Index is based on the depth to the water table, plus the depth to the top of the first significant aquifer. If the target aquifer was deemed confined, then the depth to the water table was set to the depth of the first significant aquifer, or target aquifer (based on the assumption that a contaminant particle must travel to the aquifer, not the water table, before causing an impact). 5. Mapping Calculation Process Vulnerability indices were prepared using Viewlog and Visual Basic for Applications (both part of a Microsoft Access database) for each well. The indices were interpolated using Viewlog to produce the maps. The maps produced by the process were conservative as they were based on the first significant aquifer rather then deeper production aquifers commonly used by domestic wells. Using the first significant aquifer gave consideration to the ecosystem functions of the shallow groundwater regime in the vicinity of the well. As numerous intermediate values were required by the calculations, a new table was added to the database (referred to as the HydroProp table) with the following fields: Confined flags if aquifer is confined or unconfined Depth to aquifer depth to top of first significant aquifer Quaternary Geology numeric code for quaternary geology at the location of the well Water Table interpolated water table elevation at each well GwIS calculated aquifer vulnerability index The following is a step-by-step description of the process used for calculating the vulnerability index at each well used in the map. 5.1 Data Preparation Several updates and filters were applied as the first step in an effort to improve the quality of the well data. 2 An aquitard is geologic material with significantly lower permeability and capability to transmit water than the aquifer being considered.

5 5 Steps 1. Several of the data preparation steps and the final map presentation were based on a Cartesian grid over the test area. A grid cell size of 200 m by 200 m was used for the analysis, and a 50 m grid used for the final presentation. 2. The MOE well database includes confidence indicators for the position and elevation of each well. Wells with a position confidence of within 300 m were selected for the mapping. 3. MNR provided digital elevation maps for the study area, based on 10 m contour lines and a 30 m cell size. These maps were imported and re-sampled on the 200 m project grid. The ground surface at each well was then imported from the MNR digital elevation map into the database and used as the ground surface elevation for the well. All related elevations in the database were updated accordingly. Where the difference in elevation between the original database and the digital elevation map was greater than 10 m, the well was discarded on the assumption that the well position was also likely in error. 5.2 Water Table Calculation Depth to water table is an important variable in the determination of aquifer vulnerability. To this end, a water table surface was developed from the static water levels provided in the MOE Well Log database by filtering shallow wells considered to represent water table conditions and not the piezometric head conditions in a deeper and possibly confined aquifer. Finally, as with the digital elevations, the inferred water table elevation value was written back to the database, for all wells, for use in further calculations. Steps 1. Interpolated water table surface from wells less then 20 m deep (This value was based on a review of the water level data and several cross sections of the ORM. Should better geological definition be available, the water table should be selected from wells screened in the uppermost aquifer.) 2. Compared the interpolated water table surface to the digital elevation map surface. Where the water table surface was above ground surface, the water table was corrected to be 1 m below ground surface. 3. At each well, the water table elevation from the grid cell hosting the well was retrieved and copied into the database and applied to the well as follows: a) Assigned interpolated water table value to wells without a static water level in the log. b) Kept the original water table value for wells with an existing static water level higher than the interpolated water table. c) Assigned the interpolated water table value to wells with an existing static water level less than the interpolated water table. These rules were applied to generate a conservative estimate of water table depth at all wells, and were based on the recognition that the reported static in the well log database may be artificially low if natural static had not been reached, or if downward hydraulic gradients were present. In addition, the interpolation process for the water table calculated a single average static for the cell based on distance weighting of available static levels in the cell from the cell centre. The calculated average may, therefore, not have been representative across the entire cell, and to

6 6 ensure the water table was not lowered by the interpolation, the shallowest static levels were maintained. Upon the completion of this step, the HydroProp table field called Water Table was filled with the interpolated water table elevations at each well. 5.3 Geological Updates Geological updates were necessary to: Improve the geological descriptions of each well Build a corresponding hydro-stratigraphic model for each well Steps 1. Updated geological descriptions in the MOE well database with the GSC geo-materials conversion. This was achieved using numerical code provided by the GSC. Grouped surficial geologic units at each well, less then 1 m thick, to create a single unit, at least 1 metre thick. This was achieved using a code that merged consecutive thin surficial layers into a single unit. 2. Updated the geology of the uppermost unit in the well to match the geology defined on the quaternary geology map provided by the GSC for the ORM project. 3. Prepared a hydro-stratigraphic model at each well. This involved flagging each geologic unit in the well as either aquifer or aquitard, and grouping consecutive aquifers / aquitards to produce an alternating sequence of aquifers and aquitards at each well. Table B was used to identify the aquifer / aquitard classification of the geologic units in each well. This was necessary to address cases where consecutive sequences of sand units where present, from which the first significant aquifer must be identified for the groundwater sensitivity analysis. This was achieved using code that stepped through the geological sequence of each well, assigning the aquifer / aquitard conditions and grouping consecutive aquifer layers. If no aquifers were found in the well, based on Table B, the GSC silt, silty clay, silty sand unit was re-assigned to aquifer from aquitard, and the well was reassessed for aquifers. If this switch failed again to identify any aquifers, the base of the well was set to an aquifer, on the assumption that the driller did not log the final aquifer. Upon the completion of this step, a second table, called HydroStrat was created holding a simplified geological model of each well, and the properties of the first layer (often a collection of several thin surficial layers) were taken from the quaternary geology map. 5.4 Identification of First Significant Aquifer This step used the hydrostratigraphic model developed in the previous section to identify the first significant aquifer, and determined if this aquifer was confined or unconfined. In unconfined aquifers, the vulnerability was calculated from ground surface to the water table. In confined aquifers, it was calculated to the top of the aquifer (on the assumption that contaminants must travel to the aquifer, and not the water table, before causing an impact). Aquifer selection was necessary in order to: Identify the first significant aquifer at each well Determine if the selected aquifer is confined or unconfined

7 7 Steps To select the first significant aquifer at each well, the following sequential tests were applied to the HydroStrat table. This was achieved using code that stepped down through the HydroStrat table, testing the thickness and water table conditions of each aquifer unit. Test 1: Looking at the first (upper most) unit in the HydroStrat table, if either of the following conditions were true, then the first significant aquifer was treated as unconfined, and the depth to the first significant aquifer was set to zero (ground surface). Aquifer material at surface, greater than 2 m thick, partially saturated Aquifer material at surface, greater than 2 m thick, water table no more than 3 m below (to allow for seasonal fluctuations) Test 2: If neither of the above conditions were true, the hydro-stratigraphic units of the well were examined, looking for the first occurrence of the following. If encountered, the first significant aquifer was treated as unconfined, and the depth to the top of the aquifer was set to zero. This was a conservative approach designed to discount near-surface thin fractured clays overlying aquifer material. First aquifer unit greater than 2 m thick, partially saturated Test 3: Continued stepping down the geological sequences of the well, looking for the following. If encountered, the first significant aquifer was treated as confined, and the depth to the first significant aquifer was set to the top of the first aquifer layer that met the criteria. First aquifer unit greater than 2 m thick, fully saturated First aquifer that is fully saturated (regardless of thickness) If no aquifer is detected, assume the aquifer is at the base of the well, and the top of aquifer depth is set to the depth of the well Upon the completion of this step, the HydroProp table has been updated to include, at each well, the depth to the first significant aquifer (presented as an elevation) and a flag indicating the confined condition of the selected aquifer. 6. Calculate the Geology Sensitivity Index Depth Index Calculation The Depth Index is based on the depth to the water table, plus the depth to the top of the first significant aquifer. If the well was deemed confined, then the depth to the water table was set to the depth of the first significant aquifer. Using the Ground Surface Elevation, Depth to Aquifer and Water Table Elevation and the Confined Flag from the Hydrostat Table, the Depth Index was calculated at each well and the value was stored in the HydroProp table.

8 8 Geology Index Calculation The Geology Index was calculated by stepping through the geological material of each well, from ground surface to the top of the selected aquifer (if confined) or to the water table (if unconfined), multiplying the unit thickness by the corresponding K Value, and summing the values. A vulnerability index surface was prepared by interpolating the value at each well. The method was adapted by MOE for regional scale groundwater sensitivity studies, and was based in part on work by Stempvoort et al.(1993) Like the Depth Index, under confined conditions, the depth to the water table was taken at the depth to the top of the aquifer. Table A shows recommended K Numbers for primary geological materials. For the ORM project, this table was modified to reflect the geological conditions of the ORM based on the materials produced by the GSC geo-material conversion. These modified values are shown in Table B. 7. Determination of Thresholds Groundwater sensitivity to contamination is a relative index, based largely on empirical relationships between variables with non-consistent units. Consequently, the final map uses a relative scale to indicate areas of more or less sensitivity. In an effort to make the maps applicable to land use assessments, individual values at each well were categorized into two groups, high aquifer vulnerability and low aquifer vulnerability. The threshold between these categories was selected based on the earlier Grand River Conservation Authority mapping, where areas with a Depth Index of 10 or less were defined as highly vulnerable. To apply this threshold to the ORM, the equivalent Geology Index was inferred from the graph shown in Figure 1. The graph displays the relationship between the Depth Index and the Geology Index. From the graph, a maximum Geology Index of approximately 90 was estimated to correspond with a Depth Index of 10; consequently, a conservative threshold of 100 ( ) was used to define the areas of high vulnerability on the ORM.

9 9 30 Depth Index (based on GRCA method) Geology Index (based on MOE AVI method) Figure 1. Depth Index vs. Geology Index (GRCA = Grand River Conservation Authority; AVI = Aquifer Vulnerability Index)

10 10 Table A. Generic K-Numbers Soil Type K-number gravel weathered limestone/dolomite permeable basalt 1 sand 2 peat (organics) silty sand weathered clay (<5 m below surface) fractured igneous & metamorphic rock 3 silt limestone/dolomite 4 till (diamicton) sandstone 5 clay (unweathered marine) shale 8 unfractured igneous & metamorphic rock 9 Source: Ministry of the Environment, November Groundwater Studies 2001/2002. Technical Terms of Reference

11 11 Table B. Geological Materials of the Oak Ridges Moraine (Based on the MOE well records, and after the GSC conversion) Description K Number Aquifer clay, silty clay 6 No clay, silty clay, topsoil 6 No clay, silty clay, with muck, peat, wood frags. 6 No clay, silty clay, with rhythmic/graded bedding 6 No covered, missing, previously bored 3 No diamicton: cl to cl/si matrix 5 No diamicton: cl to cl/si with gr/sa/si/cl interbeds 5 No diamicton: cl to cl/si, stoney 5 No diamicton: cl to cl/si, topsoil 5 No diamicton: cl to cl/si, with muck, peat, wood frags. 5 No diamicton: si to sa/si matrix 5 No diamicton: si to sa/si with gr/sa/si/cl interbeds 5 No diamicton: si to sa/si with muck, peat, wood frags. 5 No diamicton: si to sa/si, stoney 5 No diamicton: si to sa/si, topsoil 5 No diamicton: si/sa to sa matrix 5 No diamicton: si/sa to sa with gr/sa/si/cl interbeds 5 No diamicton: si/sa to sa with muck, peat, wood frags. 5 No diamicton: si/sa to sa, stoney 5 No diamicton: texture unknown 5 No dolomite 2 Yes fill (incl topsoil, waste) 3 No granite (poss. bedrock, prob. boulder) No gravel, gravelly sand 1 Yes gravel, gravelly sand, topsoil 2 Yes gravel, gravelly sand, with muck, peat, wood frags. 2 Yes gravel, gravelly sand, with rhythmic/graded bedding 1 Yes interbedded limestone/shale 2 No limestone 1 Yes miscellaneous; no obvious material code 3 No organic 3 No organic, topsoil 3 No potential bedrock 3 Yes rock 3 Yes sand, silty sand 2 Yes sand, silty sand, topsoil 3 Yes sand, silty sand, with muck, peat, wood frags. 3 Yes sand, silty sand, with rhythmic/graded bedding 3 Yes sandstone 5 No shale 8 No silt, sandy silt, clayey silt 4 No silt, sandy silt, clayey silt, topsoil 4 No silt, sandy silt, clayey silt, with muck, peat, wood frags. 4 No silt, sandy silt, clayey silt, with rhythmic/graded bedding 4 No

12 12 References Maps Each of the following maps is available in full colour 1:200,000 scale maps as Geological Survey of Canada (GSC) Open Files. These files are printed on an on demand basis by the GSC. They can be ordered from the GSC Bookstore at 601 Booth St., Ottawa Brennand, T. A., A. Moore, C. Logan, F.M. Kenny, H.A.J. Russell, O.R. Sharpe, and P.J. Barnett Bedrock Topography of the Greater Toronto and Oak Ridges Moraine areas, Southern Ontario: Geological Survey of Canada, scale 1: , Open File Russell, H. A. J., C. Logan, A. Moore, F.M. Kenny, T.A. Brennand, D.R. Sharpe, and P.J. Barnett Sediment Thickness of the Greater Toronto and Oak Ridges Moraine NATMAP areas, Southern Ontario: Geological Survey of Canada, scale 1:200,000, Open File Sharpe, D.R., P.J. Barnett, T.A. Brennand, D. Finley, G. Gorrelt, H.A. Russell, and P. Stacey Surficial geology of the Greater Toronto and Oak Ridges Moraine area, Southern Ontario: Geological Survey of Canada, scale 1: , Open File Reports and Papers Aller, L., T. Bennett, J.Lehr and R. Petty DRASTIC: A Standardized System for Evaluating Groundwater Pollution Using Hydrogeological Settings, National Water Well Association for the EPA. Bekesi G., and J. McConchie. Groundwater Protection Through Vulnerability Assessment CH2MHill and Waterloo Hydrogeologic. April 4, Preliminary Evaluation of Aquifer Vulnerability, prepared for City of Ottawa. EarthFX Incorporated Mapping of Groundwater Vulnerability on the Oak Ridges Moraine. Report prepared for the Ontario Ministry of the Environment. GSC web page (Quebec City) ( Comparison of four methods (EVARISK, GOD, DRASTIC and Minnesota) Guidebook on Mapping Groundwater Vulnerability Edited by Jaroslav Vrba and Alexander Zoporozec, International Association of Hydrogeologists, Volume 16. ISBN National Academy Press Groundwater Vulnerability Assessment, Washington DC, ISBN Russell, H., T. Brennand, C. Logan and D. Sharpe. Standardization and Assessment of geological descriptions from water well records, Greater Toronto and Oak Ridges Moraine areas, Southern Ontario, Geological Survey of Canada. Stempvoort, D.V., L. Ewert and L. Wassenaar Aquifer Vulnerability Index: A GIS compatible method for groundwater vulnerability mapping, Canadian Water Resources Journal, Vol. 18, No 1.