This study assessed if a permeable reactive barrier (PRB) could be used to reduce sulphate and metal concentrations of Hogarth pit lake, a sulphite-toxic (up to 2,000 mg/L) pit lake at the former Steep Rock iron mine site in Atikokan, Ontario.  Both batch reactor and flow-through reactor experiments were performed to simulate a PRB at the bench-scale in order to assess the sulphate reducing capacity of different types of organic matter.

Batch reactor experiments were run using three different treatments to promote bacterial sulphate reduction in order to lower sulphate concentrations in water from the pit lake.  Treatment 1 contained organic matter, creek sediment (sulphate reducing bacteria source), carbonate rock (acid neutralizing agent) and glacial till (non-reactive medium).  Treatments 2 and 3 were similar to treatment 1, except that treatment 2 did not include creek sediment and treatment 3 contained molasses as a nutrient.  Treatment 1 with horse manure and wood chips as the organic source resulted in >99% reduction in sulphate concentration, combined with increases in pH and bicarbonate levels, reduced redox and decreased metal concentrations.  Bacterial sulphate reduction was also initiated with Treatment 2, although did not occur as quickly as treatment 1.  The results of treatment 3 with molasses showed that no sulphate reduction occurred in the batch reactors.  Based on these results, treatment 1 was selected for the flow-through experiment to simulate a PRB at a laboratory scale.

Flow-through reactor columns were run in duplicate and filled to create different reaction chambers that contained mixtures of treatment 1.  The most effective sulphate-reducing flow-through reactors consisted of two reaction chambers separated by silica sand, which resulted in an overall sulphate reduction average of 46% and 49%.  In comparison, all other flow-through reactors achieved a 39% reduction in sulphate concentrations.  Sulphate reducing bacteria activity was evident after three weeks with reductions in redox values and sulphate concentrations and increases in bicarbonate and pH levels.  Results of flow-through reactor 1, reduced sulphate concentrations to <300 mg/L between weeks 3 and 5, and had a gradual increase for the remainder of the experiment to around 1000 mg/L.  Results of flow through reactor 5, showed a decrease in sulphate concentration to <700 mg/L between weeks 3 and 8 before also increasing to around 1000 mg/L for the rest of the experiment.  All other reactors generally decreased to 900-1000 mg/L after 2 weeks and remained around 1000 mg/L between weeks 3 and 20.

Sulphate concentration in water from the adjacent Caland pit lake, has a sulphate concentration of <300 mg/L, and a previous study at the site concluded that a treatment system which consisted of a PRB flowing into a constructed wetland has the potential to reduce elevated sulphate levels in Hogarth pit lake.  However, the flow-through experiments show that the residence time is a limiting factor in the life span of a PRB.  Also, it is possible that sulphide precipitation is limited by the availability of divalent metals, in particular Fe²⁺.

A copy of the thesis can be downloaded here

The Angel Hill Gold Zone of the West Cedartree Gold Project has been exposed over 300 meters along the internal contact of ultramafic and gabbroic rocks within the Kakagi Sill. The sill lies within the Kakagi-Rowan Lakes greenstone belt and has intruded the mafic metavolcanic rocks of the Snake Bay Formation and the felsic to intermediate metavolcanic rocks of the Emm Bay Formation. The Snake Bay Formation consists of pillowed and massive basalts and fine-grained tuffs. The Emm Bay Formation stratigraphically overlies the Snake Bay Formation and consists of tuffs and tuff breccias. Trace element geochemistry and primitive mantle normalised plots suggest that both the rocks of the Kakagi Sill and the Snake Bay Formation are comprised of metasomatsed island arc tholeiite basalts whereas the rocks of the Emm Bay Formation are characterised of a back island arc affinity.

The Angel Hill Gold Zone occurs as extensive carbonate and quartz flooding within a zone of brittle deformation termed the master fault. A distinguishing feature of the gold zone is the presence of extensive fuchsite alteration forming large mats up to 20 meters wide. Mineralization in the Angel Hill Gold Zone consists of free gold and sylvanite occurring as inclusions within and rimming pyrite grains, as well as precipitated along small fractures in quartz. Associated mineralization includes pyrite as well as galena and sphalerite. Textural relationships and stable isotope analyses of carbonate (δ¹⁸O 13.6 to 17.9 €), quartz (δ¹⁸O 8.8 to 13.1 €), pyrite (δ³⁴S -0.8 to 1.0 €), sphalerite (δ³⁴S 1.3 to 5.4 €), and galena (δ³⁴S 5.9 to 7.3 €) suggest multiple episodes of brittle deformation and fluid influx within a protracted lode gold system. The combination alteration assemblages and stable isotope characteristics suggest that the Angel Hill Gold Zone formed as a result of listwanite metasomatism with input of fluids from both magmatic and metamorphic sources. These fluids propagated as a result of the brittle faulting that occurred due to a competency contrast between the ultramafic footwall and the gabbro hangingwall during regional deformation.

A copy of the thesis can be downloaded here

Pit lakes are often a planned part of an open pit mine closure where the excavations are expected to flood and water quality is not an issue. Common environmental issues regarding pit lakes include their rebound rate, hydrodynamic behaviour and water quality. The water quality of pit lakes can be influenced by their hydrodynamics, for example overturn in a holomictic lake can transport dissolved oxygen down to submerged tailing resulting in the production of acid mine waters if sulphide minerals are present, or the unexpected overturn of a meromictic pit lake can bring stagnant, dissolved metal laden waters to surface that may be toxic to aquatic life. Where water quality is of concern and pit lakes outflow into adjacent watersheds their behaviour can determine if noxious material will be brought to the surface and released. At the former Steep Rock Iron Mines property near Atikokan, Ontario, three pit lakes are currently flooding and will eventually join to form a super pit lake before they outflow into the West Arm and subsequently Seine River system. Previous studies on two of the pit lakes, Caland and Hogarth, have shown that the pit lakes are meromictic and holomictic, respectively, and that both have elevated sulphate concentrations. The aim of this research was to: i) evaluate existing rebound models by modeling rebound and assessing which parameters exert the greatest influence on the rebound rate; and, ii) develop hydrodynamic models of Caland and Hogarth pit lakes to assess if their current limnology will change as rebound continues and they outflow into the West Arm.

Rebound models are constructed using two approaches and compared to the Ontario Ministry of Natural Resources Regional Engineering model that accurately predicted water levels to 2011. The first rebound modeling approach uses two curves to model the stage-volume relationships, a hypsometric curve and a surface area versus elevation curve. The second approach fits an exponential curve to measured water elevations and then future water elevations are forecasted by extrapolation. Rebound Model 2B constructed following the first approach matched measured water elevations best for the two pit lakes and predicts 2010 measured water elevations better than the Regional Engineering model. Model 2B predicts that Caland will flow into Hogarth in 2070 and that the new Steep Rock pit lake will outflow into the West Arm in 2087, 18 years longer than predictions made by the Regional Engineering model. Based on the water balance parameter sensitivity analysis, the difference between this study's predictions and those of the Regional Engineering model is the result of different pit volume calculation methods. In this study's rebound models the stage-volume relationships for Hogarth are more accurate than for Caland, suggesting that in future work, at minimum, linear interpolation should be used to define the volume in Caland pit lake.

This study is the first to model the hydrodynamics of Caland and Hogarth pit lakes. The Dynamic Reservoir Simulation Model (DYRESM) was used to: i) assess if it can accurately model the current pit lake conditions; and, ii) model the future conditions in Caland and Hogarth for when the pit lakes join and when they outflow to the West Arm. The model salinities are discussed to assess the future toxicity of the pit lakes. DYRESM simulations of current conditions accurately portray the observed limnological characteristics of Caland and Hogarth pit lakes, including: i) that Caland is meromictic and has a lower salinity relative to Hogarth; and, ii) that Hogarth develops a temporary meromix. Simulations of when the two pits join indicate that the freshwater lens in Caland will be maintained, but is thinner, and that Hogarth develops a meromix, which is maintained throughout the simulations. Simulations of when the pit lakes outflow into the West Arm indicate that Caland will maintained its upper freshwater lens and that a fresh water lens is only briefly present in Hogarth. In most cases, variations of the simulations for current and future pit lake conditions, including additional inflows, alteration of the inflow salinities, and the use of a slower rebound rate to define the DYRESM water balance, only produced minor changes in the simulation.

A linear trend between sulphate concentrations and salinity exists for water samples from Caland and Hogarth. Based on this trend, the DYRESM salinity profiles suggest that the waters that outflow from Caland into Hogarth will have sulphate concentrations ranging from 0 mg/L to 100 mg/L and that waters that outflow from Hogarth will have sulphate concentrations ranging from 1700 mg/L to 1900 mg/L. In general, the sulphate concentrations in Caland are below maximum acceptable limit of all water quality standards while those in Hogarth exceed all water quality standards. These results suggest that the waters that outflow from the pit lakes will be toxic.

DYRESM can be used to simulate the future hydrodynamics of Caland and Hogarth pit lakes, however, future studies and field investigations should address some of the areas of uncertainty in the DYRESM simulations for Caland and Hogarth pit lakes, including constraining seep and groundwater volumes and chemistry, on site meteorological monitoring and measurement of the light extinction coefficient.

A copy of the thesis can be downloaded here

The Hamlin Lake area is located approximately 120km southwest of Thunder Bay, Ontario, in the Shebandowan Greenstone Belt of the Wawa Subprovince in the Superior Province. The area has been explored for its copper and gold mineralization for more than 50 years, but has only recently been treated as aniron-oxide copper gold (IOCG) occurrence. The aim of this study was to characterize the alteration at the occurrence through space and time, and to relate this paragenesis to the formation of mineralization. Localized field mapping and re-logging of drill core established several relationships between styles of alteration. Rocks were examined in hand specimen and thin section, and minerals were examined with scanning electron microscope (SEM) to identify mineral assemblages and associations. Consistent variations in space could not be mapped, so textural relationships were used to relate the timing between six styles of alteration and the local brecciation that hosts the potential ore. These are: (1) sodic, (2) early potassic, (3) calcic(-iron), (4) late potassic, (5) carbonate, and (6) silicic.

A copy of the thesis can be downloaded here.

A geoarchaeological investigation was north of Highway 11/17, 34km east of Thunder Bay, Ontario.  Five archaeological sites (Mackenzie 1, Mackenzie 2, RLF, Woodpecker 1, and Woodpecker 2) and seven additional sediment exposures were examined for stratigraphic analysis to accompany the archaeological excavations.  River-mouth sediments at 268m asl and a series of deltas indicate that the study area was subaqueous while placement of the Superior lobe prevented drainage to the Superior basin.  This elevation is consistent with Lake Beaver Bay, an ice-contact lake that received glacial meltwater from the north (the Hudson Bay lobe) as well as the south (the Superior lobe).  This is demonstrated by southward and northward prograding deltaic sequences within the study area.  

As the Superior lobe made its final retreat, Lake Beaver Bay dissipated into the Superior basin marking the beginning of the Minong phase, likely around 9,900 ¹⁴C yrs BP.  Additional sequences representing river-mouth, beach shoreface, and deltaic depositional environments indicate that a series of shorelines within the study area represent subsequent Minong lake levels.  The highest, and likely oldest of these strandlines is an erosional feature at 256m asl, consistent with wave-cut terraces previously identified in the Thunder Bay region.

Relative lake level drops occurred, likely due to a combination of gradual erosion of the Nadoway Point sill and isostatic rebound of the recently deglaciated land.  Beach and river-mouth sequences representing subsequent shorelines are located at 249m, 243m, and 240m asl.  Artifacts on each of these beach terraces suggest they were occupied by Paleoindian groups.

The occupation layer(s) at the Mackenzie 1 site are strongly bioturbated, although the sediment matrix is consistent with underlying beach sediments in the north and river-mouth sequences in the south.  The site is about 10,000 m², and 378 Paleoindian projectile points were recovered along with additional bifaces, other formal and expedient tools, as well as lithic debitage.  The frequency of artifacts and site size likely indicate that Mackenzie 1 was successively occupied over an extended time period of time.  However, absence of an unconformity separating the visible stratigraphy from the massive occupation layer(s) may indicate that the site was inhabited soon after deposition ceased.  This likely places site occupation within the Minong phase (dating to ~10,500 to 9,000 cal BP). 

Artifacts recovered from the RLF archaeological site are also within a bioturbated sediment matrix consistent with underlying stratigraphy.  Lithofacies indicate that soon after the beach shoreface sediments were deposited, the beach terrace was utilized by mobile Paleoindian groups.

A shoreline at 240m asl is evidenced by a wave-cut feature and beach sediments at the Woodpecker sites, river-mouth sequences at the Mackenzie 2 site, and beach shoreface deposits at a roadcut exposure.  Presence of artifacts and charcoal within beach sediments at the Woodpecker 2 site provides evidence that occupation was contemporaneous with active beach formation.  However, the majority of recovered artifacts at Woodpecker 1 and Woodpecker 2 are associated with bioturbated sediments consistent with underlying stratigraphy.  Most likely, the Woodpecker sites were occupied along an active Lake Minong margin, and subsequently inhabited soon after the relative lake level dropped again.  The artifact matrix at the Mackenzie 2 site similarly suggests that occupation occurred soon after deposition of the underlying river-mouth sequences

Two additional exposures revealing a deltaic sequence and beach sediments suggest that the relative lake level lowered to 233m, and subsequently to 224m asl.  This lowest shoreline identified within the study area likely represents the beginning of the Post-Minong phase.

All five archaeological sites are strategically placed on beach terraces, which is consistent with most presently known Paleoindian habitations in Northwestern Ontario.  As well, The Mackenzie and Woodpecker sites likely had access to a river, making them ideal for fishing as well as hunting at the river crossings.  The study area provides additional evidence that lake margins and river-mouths were highly attractive campsites for mobile Paleoindian groups.  In addition, artifacts recovered from within beach sediments at Woodpecker 2 suggest that the Thunder Bay region was first occupied soon after deglaciation.  The Mackenzie, RLF, and Woodpecker sites were likely inhabited between about 9,900 and 9,000 ¹⁴C yrs BP. 

A copy of the thesis can be downloaded here

Musselwhite Mine and Hammond Reef are shear-zone-hosted gold deposits located in Northwestern Ontario, in the Western Superior Province of the Canadian Shield. A detailed microscopic investigation of three gold-hosting lithologies from Musselwhite Mine and Hammond Reef demonstrate close similarities in the microstructures which host gold and the relative timing of gold mineralization. The gold deposit at Musselwhite Mine is hosted by metamorphosed banded iron formation, while Hammond Reef is hosted by metamorphosed tonalite. Despite the difference in rock types, Musselwhite Mine and Hammond Reef are similar in that that they are located proximal to regional shear zones, have undergone regional metamorphism and are dominated by ductile deformation. Although these two deposits are hosted by completely different lithologies the microstructures which host gold are very similar, indicating structural control on mineralization at a microscopic scale.

The most common gold-hosting microstructures in these gold deposits result from heterogeneous deformation. Gold mineralization commonly occurs in fractures in competent minerals such as garnet and pyrite in every lithology in this study. These gold-hosting fractures do not extend throughout the matrix but are restricted to the competent minerals because the competent minerals are resistant to the ductile deformation around them. Gold mineralization is also associated with other deformation-induced microstructures, such as strain shadows. Gold mineralization is hosted by metamorphic minerals at both Musselwhite Mine and Hammond Reef. Gold inclusions occur at Musselwhite Mine in such metamorphic minerals as grunerite and garnet and at Hammond Reef gold inclusions occur in metamorphic muscovite. Also gold mineralization commonly occurs on plane defects, for example on grain boundaries.

Relative timing of gold mineralization is shown to have occurred during ongoing metamorphism and deformation. Inclusions within metamorphic minerals indicate that gold mineralization must have occurred before or during metamorphism, while gold mineralization associated with deformational features indicate gold mineralization to have occurred during or after deformation. This thesis demonstrates close similarities between the relative timing of gold mineralization as well as the microstructures which host gold at Musselwhite Mine and Hammond Reef.

Maura is currently working as a geologist for Fladgate Exploration Consulting Corporation in Thunder Bay.

A copy of the thesis can be downloaded here

Banded iron - and manganese - rich precipitates were collected from the lake bottoms of Lake Charlotte (Nova Scotia), Lake Shebandowan (Ontario) and Sowden Lake (Ontario). Investigations of study areas at the macro, meso and micro scale were conducted to understand the iron-manganese rich- nodules in their natural environment. The nodules appear as circular precipitate bands which alternate between high concentrations of iron and manganese.  Analysis of precipitates revealed that those from Lake Charlotte are highly concentrated in arsenic. Lake Shebadowan and Sowden Lake samples are highly concentrated in phosphorous.

Correlation between iron, arsenic and phosphorous suggests oxidation and precipitation of these elements in the same bands of the nodule. Iron relies on the Eh and pH of an environment to precipitate from solution. At a redox boundary in a near neutral environment, iron is able to oxidize as a sediment coating and co-precipitate arsenic and phosphorous from the water.  An affiliation between manganese, barium and cobalt suggests precipitation of these elements in the alternate bands present in the nodule. Barium and cobalt are able to co-precipitate with manganese by either penetrating a manganese oxide by means of protonation, or oxidize and become interchangeable with Mn⁴⁺. 

The growth of the nodules at Shebandowan and the majority of Lake Charlotte sites were probably affected by a redox boundary created by the diffuse upward flow of groundwater with lower Eh than the oxidized lake water. It is likely that photosynthetic and iron and manganese oxidizing microorganisms are present in a bacterial mat covering the nodules and probably played a role in their precipitation. Analysis of the growth mechanisms of precipitates revealed in Sowden Lake and the Granite Islands site of Lake Charlotte were inconclusive.

Research involving iron and manganese rich precipitates may be used for environmental implications involving heavy metals and arsenic. Examples of remediation areas that would benefit from precipitate research include drinking and waste water filtration and reducing the environmental footprint of mine tailing waste sites.

A copy of the thesis can be downloaded here

The McFaulds Lake area (commonly known as the Ring of Fire), has been the site of much recent exploration within northern Ontario.  The area represents a recently discovered Archean greenstone belt which is host to world class chromite deposits along with significant Cu-Zn VMS, magmatic Ni-Cu-PGE and Fe-Ti-V occurrences.  Much emphasis has been placed on the chromite mineralized ultramafic intrusions with little attention focused on the Fe-Ti-V mineralized ferrogabbroic intrusions.

The Butler and Thunderbird intrusions represent the best described intrusions within the volumetrically significant ferrogabbroic suite within the McFaulds Lake area.  These intrusions are characterized by a suite of well layered magnetite-ilmenite rich rocks which are dominantly composed of gabbroic to anorthositic units with lesser stratigraphically conformable units composed of pure magnetite-ilmenite.  The Fe-Ti oxide rich layers contain variable vanadium mineralization and low chromium contents within magnetite (up to 2.45 V₂O₅ wt. %, 0.99 % Cr₂O₃ wt. %) and ilmenite (up to 0.57 V₂I₅ wt. %).  The massive and semi-massive oxide layers occur as basal members of repeated cycles characterized by sharp lower contacts which grade upwards into oxide-rich pyroxenite, followed by oxide-bearing leucogabbros and/or anorthosites.  The layers are believed to be cause dominantly by magmatic convection currents within a system which is at least partially open to oxygen.  No evidence has been found to suggest multiple pulses of magma.  Oxide-silicate liquid immiscibility is thought to only occur within the evolved, apatite-bearing margins of the Thunderbird intrusions; however, additional drilling may reveal further apatite mineralization.

The ferrogabbroic intrusions are thought to have originated from a shallow depleted mantle source, possibly related to a plume event.  The ferrogabbros have likely undergone a two stage differentiation to account for the extreme iron enrichments.  The first stage is characterized by an anhydrous, tholeiitic melt, within the upper mantle (above the garnet stability field, <110km) which underwent Fe-Ti enrichment due to the crystallization of Fe-poor phases (e.g., olivine, plagioclase, etc.) within a system closed to oxygen.  The second stage is considered to be a very shallow intrusion within the McFaulds Lake mafic-felsic volcanic rocks.  This final stage is characterized by a system which was at least partially open to oxygen from an originally reduced magma (<QFM buffer).  These magmas initially crystallized Cr-V-rich magnetite-ilmenite horizons and gradationally evolved into Cr-V-poor, apatite-bearing ferrogabbros.  These ferrogabbros likely share a parental magma with the coeval Cr-Ni-PGE-bearing ultramafic intrusions of the McFaulds Lake greenstone belt.  Additionally, spatial and geochronological evidence suggests that abundant VMS-style mineralization within the McFaulds Lake area may be a result of a thinned lithosphere during plume tectonics.

A copy of the thesis can be downloaded here

Glacial sedimentary rocks of the Huronian Supergroup crop out along the north shore of Lake Huron and were likely deposited on what is thought to have been a divergent continental margin (Fralick and Miall, 1981; 1989). The rocks of the Gowganda Formation record one of three glacial events preserved in the Supergroup and are therefore of interest for developing further glaciomarine models (Puffett, 1974) and furthering understanding of this early stage in Earth's history. Data has been collected in five main study areas in an attempt to cover as much of the ancient continental margin as possible.  The study areas include Espanola, Elliot Lake, Thessalon and Cobalt, Ontario and Marquette, Michigan. There are two glaciogenic formations in the Marquette area of Paleoproterozoic age, the Reany Creek Formation and the Enchantment Lake Formation.  The Enchantment Lake Formation has been chronostratigraphically correlated to the Huronian Supergroup based on U-Pb age determination on detrital zircon, 2317±6 Ma, and diagenetic xenotime, 2133±11 Ma (Vallini et al., 2006).  As these formations are present in such close proximity to each other, and there are no Archean glacial events recorded in the rest of the Canadian Shield, it is reasonable to correlate them with the Gowganda Formation, the thickest and most commonly preserved of the three Huronian glacial events.

Stratigraphic sections were compiled in each of the study areas and the sedimentary rocks were grouped into seven lithofacies associations (LA): 1) Planar Cross-Stratified Sandstone LA, 2) Basal Breccia LA, 3) Diamictite LA, 4) Interlayered Siltstone and Fine-Grained Sandstone LA, 5) Slump LA, 6) Heterogeneous Sandstone LA and 7) Quartz-Rich Sandstone LA.  These lithofacies associations likely represent a sequence of depositional environments on a shallow continental shelf.  Initially, the shelf was dominated by what were likely large-scale, low-angle sandwaves, interbedded with successions of wavy bedding and possible hummocky cross-stratification indicating an open-water setting with tidal and storm processes reworking the sediments.  The shelf then gradually evolved into an environment dominated by diamicite layers.  The diamictite layers have dropstones as well as evidence of current activity indicating outsized clasts were likely being introduced into the environment as ice-rafted debris.  Resedimentation events in the form of debris flows are thought to account for conglomeratic layers that are common in the Diamictite LA.  The Interlayered Siltstone and Fine-Grained Sandstone LA, along with the Slump LA, seem to indicate deposition from suspension in a prodelta setting where large slump events are common.  The gradual transition into a more sandstone dominated LA, with an abundance of current-related sedimentary structures, is indicative of the shallowing and coarsening upwards succession common to deltaic deposits.  A final transition into the Quartz-Rich Sandstone LA indicates a return to a sandy, current-dominated open continental shelf environment with abundant tidally generated sedimentary structures such as herringbone cross-stratification.  The Cobalt study area differs from this overall model in that evidence of grounded ice is present.  Less exposure in the Marquette study area makes it difficult to draw overall conclusions on the evolution of the continental shelf but deposition in a subaqueous glacial outwash fan is hypothesized.

A copy of the thesis can be downloaded here

Dyke rocks occurring in the Coldwell alkaline intrusive complex, N.W. Ontario, were studied by various petrographic and geochemical means.  The resulting observations indicate that 5 types of mafic alkaline lamprophyres, distinguished on the basis of chemistry and mineralogy according to the scheme of rock (1977), and 1 type of felsic tinguaite occur in the complex (in approximate order of abundance):