NI Technical Report: Resource Estimation Sheps Lake and Perault Lake Properties Labrador, Canada

Transcription

1 NI Technical Report: Resource Estimation Sheps Lake and Perault Lake Properties Labrador, Canada Respectfully submitted to: New Millennium Iron Corp. Date Submitted: March 19 th, 2013 Effective Date: January 17 th, 2013 Prepared By: Maxime Dupéré P.Geo. Minerals Services 10 boul. de la Seigneurie Est, Suite 203, Blainville, Québec Canada, J7C3V5 t (450) f (450) Member of SGS Group (SGS SA)

2 Resource Estimation of the Sheps and Perault Taconite Properties Page ii Table of Contents 1 Summary Introduction General Terms of Reference Source of Information Currency, Units, abbreviations and Definitions Reliance on Other Experts Property Description and Location Location Property Description and Ownership Permits Accessibility, Climate, Local Resources, Infrastructure and Physiography Accessibility Climate Local Resources and Infrastructure Physiography History Geological Setting and Mineralization Geological Setting Regional Geology Property Geology General Lithology Structure Mineralization Exploration Drilling Sheps Lake property Perault Lake property Sample Preparation, Analyses and Security NML QA/QC Duplicate samples at MRC Data Verification SGS independent sampling and assaying Mineral Processing and Metallurgical Testing Mineral Resource Estimates... 42

3 Resource Estimation of the Sheps and Perault Taconite Properties Page iii 13.1 Drill Hole and Sample Data Outside Limits Inside Limits Compositing, and Statistical Analysis Block Grade Interpolation Block Categorization and Mineral Inventory Estimated Resources Adjacent Properties Other Relevant Data and Information Interpretation and Conclusions Recommendations References Certificate of Qualified Person... 82

4 Resource Estimation of the Sheps and Perault Taconite Properties Page iv List of tables Table 2.1: List of Abbreviations Table 4.1: Summary of NML mineral claims holding at Sheps Lake & Perault Lake Table 7.1: Regional Stratigraphic Column Table 7.2: Detailed taconite stratigraphic column for the Property area Table 10.1 Statistics of NML duplicated data Table 11.1: Statistics of SGS check sample data Table 13.1: Statistics of litho intervals in drill holes Table 13.2: Summary of SG Regression Table 13.3: Volumetric of the litho units in the resource model Table 13.4: Statistics of sample data according to litho units Table 13.5: Statistics of composites according to lithological units Table 13.6: Summary of blocks classified by pass Table 13.7: Estimation settings search parameters Table 13.8: Estimation settings ellipse parameters Table 13.9: Statistics of interpolated block values according to litho units Table 13.10: Mineral Inventory of Sheps and Perault at various weight recovery Cut-Offs Table 13.11: Estimated Resources of Sheps & Perault by Unit and Total Table 13.12: Estimated Resources of Sheps by Unit and Total Table 13.13: Estimated Resources of Perault by Unit and Total Table 16.1: Estimated Resources for the Sheps Lake & Perault Lake Properties... 78

5 Resource Estimation of the Sheps and Perault Taconite Properties Page v List of Figures Figure 4.1: Location map of the Sheps Lake & Perault Lake properties Figure 4.2: Sheps Lake & Perault Lake Properties & Others Owners Figure 7.1: Geological map of the Labrador Through Figure 7.2: Sheps Lake and Perault Lake geology Figure 8.1: Taconite targets on the residual magnetic grid in the Sheps Lake and Perault Lake area. 30 Figure 9.1: Drilling program and sections lines of the Sheps Lake area Figure 9.2: Drilling program, section lines and geology of the Perault Lake area Figure 10.1: Correlation plots of results from NML duplicates Figure 11.1: Correlation plots of results from SGS check samples Figure 13.1: Map of Sheps drill holes Figure 13.2: Map of Perault drill holes Figure 13.3: Derivation of lateral limits for Sheps resource model Figure 13.4: Derivation of lateral limits for Perault resource model Figure 13.5: Overburden thickness for Sheps-Perault resource model Figure 13.6: Thickness of mineralized units from modeled contact surfaces Figure 13.7: Histogram of % Fe Head for composites of various lithological units Figure 13.8: Histogram of %Wt Rec Head for composites of various lithological units Figure 13.9: Histogram of %Fe concentrate for composites of various lithological units Figure 13.10: Histogram of %SiO 2 Concentrate for composites of various lithological units Figure 13.11: Correlation of 6 m Composite values in the LC unit Figure 13.12: Block estimates in a bench for LRGC unit Figure 13.13: Drill Sections through Block Model Figure 13.14: Mineral Inventory with Weight Recovery Cut-Off Figure 13.15: Perspective view of optimized pit shell Figure 14.1: Map of adjacent Properties... 75

6 Resource Estimation of the Sheps and Perault Taconite Properties Page 1 1 Summary The SGS office (SGS Geostat) in Blainville (QC), part of ( SGS ) was commissioned by New Millennium Iron Corp. ( NML ) on July 25 th, 2012 to prepare an independent estimate of the mineral resources of their Sheps Lake & Perault Lake taconite properties ( Properties or Property ) in Labrador (NL) close to the northern Québec (QC) provincial boundary. The mineral resource estimate was completed by SGS based on data available from drilling data completed in the spring of 2011 and The mineral resource estimate was completed in accordance with National Instrument : Standards and Disclosure for Mineral Projects. This report represents the first NI compliant resource estimation on the Sheps Lake & Perault Lake properties. This report on the mineral resource estimation at the Property was prepared by Maxime Dupéré, P. Geo. Mr. Dupéré is responsible for the data quality sections, the geological modelling and resource estimation. Mr. Dupéré visited the Property between August 20 th to 22 nd, 2012 for a review of exploration methodology, sampling procedures and; to conduct an independent check sampling of selected mineralized drill intervals. Property description and location The Sheps Lake and Perault Lake Properties are located approximately 20 km and 17 km southwest of Schefferville, QC. The Properties are situated in unorganized territory. The properties consist of a total of 17 contiguous mineral rights licences totalling 506 claims and 25 km². The claims have not been marked out on the ground or legally surveyed, nor is there any requirement to do so. NML detains 100% of the Perault property. NML detains 80% of interests and Naskapi Nation of Kawawachikamach (NNK) 20% of the Sheps property. Accessibility, Climate, Local Resources, Infrastructure and Physiography Access to exploration drilling area was by truck and snowmobile. A few lakes around the Properties are accessible from Schefferville via chartered fixed-wing float aircraft and such aircraft was used to support NML s 2011 and 2012 drilling programs along with a helicopter for drill and crew moves. The Sheps Lake & Perault Lake area has a sub-arctic climate. There is ample room available on the Sheps Lake & Perault Lake Properties for the establishment of mining and processing operations, waste piles and a tailings management area. Topography is flat to gently rolling, with the occasional more precipitous area. The area is well drained, has a few swampy areas and is covered by sparse northern boreal forest consisting of stunted spruce, alders and willows. Regional geology The Sheps and Perault properties are located in the Churchill Province, of the Labrador Trough ("Trough") adjacent to Archean basement gneiss. The Trough comprises a sequence of Proterozoic sedimentary rocks, including iron formation, volcanic rocks and mafic intrusions. NML s Sheps Lake & Perault Lake properties are located north of the Grenville Front in the Churchill Province where the Trough rocks have been only subject to greenschist or sub-greenschist grade

7 Resource Estimation of the Sheps and Perault Taconite Properties Page 2 metamorphism. The principal iron formation unit is known as the Sokoman Formation. Iron formations in this part of the Trough are taconites, which are weakly metamorphosed. Property geology and Mineralogy The taconite iron formation in the Sheps Lake & Perault Lake area is part of the Sokoman Formation occurring at the western margin of the trough. The taconite is well exposed as a long linear belt 1.5 to 2.0 km wide for a total length of approximately 6 km in a northwest-southeast trend which constitutes the AM anomalous zone. Within the Sheps Lake and Perault Lake properties the structure is very simple, the iron formation is generally northwest-southeast striking and dipping 5⁰ to 10⁰ to northeast. Within oxide iron formation units, the most notable feature in mineral composition is the rather abrupt changes from dominantly magnetite (LC) to dominantly hematite (URC, JSP), and corresponding change of the silica from chert over to jasper. These oxidation potential variations and changes in iron grade define the member lithology units. The occurrence of iron carbonate minerals, principally siderite, ferrodolomite and ankerite are widespread but are more abundant in the upper (LC) and lower (LRGC) units. Iron silicates, minnesotite and stilpnomelane occur in LC, GC and LRGC units. These features all appear to be related to primary deposition. Exploration During 2010 New Millennium Iron Corp ( NML ) carried out Airborne High Resolution Magnetic Survey ( AM ) over the Millennium Iron Range ( MIR ) to identify additional, potential magnetic taconite deposits similar to the proven LabMag deposit in the Howells River Area and the KéMag deposit in Québec. The survey covered the taconite belt over a length of 150 km starting south of Perault Lake, to north of Lac Ritchie in Qubec. The magnetic taconite formation of the MIR occurs as a linear strip approximately 4 km wide, starting west of Menihek Lake and extending northwest to Lac Otelnuk in Québec over a length of 210 km. The Airborne survey conducted by Fugro covers all the licences and claims west of longitude 67⁰W in NL and QC. Several strong magnetic anomalies were outlined by Jean Hubert, Geophysical Consultant for NML, who interpreted the survey data. Drilling Based on the presence of strong magnetic anomalies identified in 2010, a detailed exploration drilling program was undertaken in the Sheps Lake and Perault Lake AM anomalies area during the spring of 2011 and Drilling at Sheps The NML drilling program started on October 21, 2011 and two holes were drilled in the Sheps Lake area totalling 158 m. The next year, the drilling program started on April 20 with 3 drills and concluded on May 30, During that period 23 holes were drilled in the Sheps Lake area totaling 1,920.1 metres in length. The drilling was carried out on a grid of lines spaced 1 km apart with 3 to 4 holes spaced 300 to 500 m on each section line.

8 Resource Estimation of the Sheps and Perault Taconite Properties Page 3 Drilling at Perault In the Perault Lake AM anomaly area, the drilling program started on March 02, 2012 and concluded on June 10, During that period 48 holes were drilled totaling 3,890.8 metres. The area covered by drilling is approximately 18 km long and 1 to 2 km wide and encompasses licences M and M. The program started with one drill and was subsequently joined by 2 additional drills provided by the contractor Downing Estate Drilling Ltd. It was decided to drill initially on lines spaced 2 km apart starting from section line 130 in the south with 2 holes spaced 300 to 400m apart on each line. This pattern was followed until the drills reached section line 200. Between sections 200 and 310, drilling was carried out on lines spaced 1 kilometre apart 3 to 4 holes spaced 200 to 300 metres along section lines. Samples and core duplicates for the 2011 and 2012 program totalled 967 samples. QAQC by NML and Data verification Over the course of the drilling and sampling program, NML has sent 49 duplicate samples to the Midland Research Center lab ( MRC ). Those samples are actually the second half of the core. At MRC, those samples were subjected to the same Davis Tube testing as the original half core for the same intervals. Duplicate samples are from 47 of the 73 holes. Multiple samples in the same hole are in holes 12PL1018D (2). Their length varies from 1.9 to 9 metres with an average of 5.4 metres. Most of duplicate samples are in the LRGC unit (15) followed by the LC (9), the JUIF, the PGC and the URC units (5 in each), the LRC unit (4), the GC unit (1) and the LIF units (3 in each). During the site visit conducted from August 24 th to 26 th,2012, SGS Geostat completed 32 analytical checks of drill core duplicate samples taken from selected NML 2012 diamond drill holes on the Sheps Lake and Perault Lake properties as well as the nearby Howells River North property (not part of this report) as part of the independent data verification program. SGS Geostat also conducted verification of the laboratories analytical certificates and validation of the data set supplied by NML for errors and discrepancies. Only a total of 10 mineralized core duplicated were collected from the Sheps Lake (5 samples) and Perault Lake (5 samples) area. Resource Estimation- General The resources reported here are derived from a computerized resource block model. The construction of that model starts with drill hole data, which serve as the basis for the definition of 3D mineralized envelopes with resources limited to the material inside those envelopes. The next step is the selection of drill hole data within the mineralized envelopes in the form of fixed length composites and then the interpolation of the grade of blocks on a regular grid and filling the mineralized envelopes from the grade of composites in the same envelopes. All the interpolated blocks below the overburden/bedrock contact surface make up the mineral inventory and they are classified according to proximity to composites and corresponding precision/confidence level. An optimized pit shell was done to verify the validity of the weight recovery (Wt Rec.) cut-off grade used and however it was not used to restrain the mineral resources and its estimate. Limits of Resource model Within top and lateral limits described in the previous section, the proposed Sheps-Perault resource model looks like a multilayered cake with layers corresponding to lithological units identified by

9 Resource Estimation of the Sheps and Perault Taconite Properties Page 4 NML geologists. Up to 7 layers have been modeled i.e. from top to bottom: LC = lean chert, JUIF = jasper upper iron formation, GC = green chert, URC = upper red chert, PGC = pink-green cherty, LRC = lower red chert and LRGC = lower red green chert. Units below LRGC (i.e. LIF, RS, BC and QTE) have not been modeled since they just appear in just a few holes and they do not show any obvious mineralization. Thus the bottom of the resource model is the interpreted footwall surface of the LRGC layer. The modeling of contact surfaces between retained layers is based on the interpreted line contacts by NML geologists on drill sections. Those lines are discretized at every 50 m and TIN faces connect points on lines of the same nature on two adjacent sections. At both south and north extremities, lines on the last drill section are extruded horizontally along the N323 azimuth. Finally modeled contact surfaces between layers are chopped by the modeled overburdenbedrock contact surface. Resource Estimation- Assay data and compositing Most of samples are in the three units LC, PGC, and LRGC however all seven units are utilized in the block grade interpolation. The %Fe of head averages about 30% in JUIF, URC, LRGC, and PGC (28.5%) but it is lower in the LC marker horizon (25%) and much lower in the GC marker horizon (18% average) and higher in the LRC unit (average 34%). Average recovery is high in LRC (29%) and JUIF (26%), moderate in LRGC (22.4%), URC (22.3%), PGC (21.4%) and LC (18.3%) and expectedly low in GC (3.9%). The quality of the magnetic concentrate is about the same in all units, even in those with very low recoveries. For samples in all zone presented, the average %Fe keeps around 69%-71% and the average %SiO 2 between 1.3% and 2%. Although the majority of original samples are 6 metres long, there exists both some shorter and some longer samples. Therefore, statistics might be biased by some high or low values measured on short or long intervals. The standardization of sample size is done by numerical compositing. The most natural composite length is 6 m, which is the composite length of the majority of samples. Statistics of composite data in the seven units of interest are consistent with statistics of sample data in the same unit. Resource Estimation- Correlations and variography Correlations between composite values are about the same in all units. There is no useable correlation (R=+0.4) between recovery and %Fe of head nor between %Fe of head and %Fe of concentrate (R=+0.2). There no correlation between %Fe of concentrate and the %weight recovery thus indicating the quality of iron concentrate has no bearing on the weight recovery. The really strong (and expected) correlation is the negative one (R= to ) between the %Fe and the %SiO2 of concentrate, except for the URC unit (R=0.287) however there is a limited range of high and low values. Resource estimation- Block interpolation The four quality parameters analyzed in the previous section are interpolated in blocks on a regular grid. That grid is oriented parallel to the drill sections i.e. columns are numbered along a rotated x axis along the N53 azimuth and rows along a rotated y axis along the N323 azimuth. The origin of the grid i.e. the center of the block in the first column (numbered from west to east) and first row (numbered from south to north) has the following UTM coordinates: 635,300 me and 6,050,650 mn. The selected block size is 25 m along the rotated x and 50m along the rotated y. That size is obviously tiny when compared to the drill hole grid of 500x1000 m. The main reason to choose blocks with such a small size is to better reflect the geometry of litho units slowly dipping to the NE.

10 Resource Estimation of the Sheps and Perault Taconite Properties Page 5 It is also the size of blocks used in the resource modeling of the LabMag and KéMag deposits further to the south which both belong to NML. Vertically, blocks are 15m high. Although the bench heights previously used for the KéMag and LabMag Pre-Feasibility studies was 13 m, the most recent work by mining consultant Met-Chem on those projects has suggested that a bench height of 15 m would be more suitable. In a first pass, the volumetric fraction is calculated in blocks of each of the seven litho units above the footwall of LRGC i.e. LC, JUIF, GC, URC, PGC, LRC, and LRGC. Then the four quality parameters of block fractions in these units are interpolated from composite data in the same unit. Interpolation is done by inverse distance squared (ISD2). The basic search ellipsoid is a flat 1200m x 600m x 50m tilted 10 to the N53 (a review of interpreted litho units on sections shows this 10 dip angle for most on them). The 1200m x 600m elliptic outline on sub-horizontal planes is designed to capture composites from at least 4 neighbor holes on the 1000m x 500m nominal grid. For example, In the first interpolation run, the minimum number of composites was set to 5 (3 in LRC, URC and the thin GC unit) in a minimum of 3 drill holes (maximum number of composites from the same drill hole is 2) within the 1200m x 600m x 50m search ellipsoid for allowing the block interpolation in that run. The maximum number of composites retained is 25 (15 in LRC, URC and the thin GC unit). 2 consecutively larger search ellipsoids (3 for GC) were used afterwards and are described in the Resource estimates section. Resource Estimation- Categorization The Sheps-Perault magnetite bearing taconite formations are currently recognized by 73 vertical drill holes on a grid of 1km by 0.5 km, 3 of which were abandoned and do not contain any assay values. Those holes cover a NW-SE area with a strong magnetic anomaly recognized by the Fugro airborne survey of This area is bounded to the NW and SE by two major fault zones easily recognizable from the survey. Aside from those two bounding fault zones, the map of the magnetic anomaly does not show any major discontinuity. The geological continuity of the mineralized units has been demonstrated by the results from the 70 holes. In most holes, with the occasional disappearance of the thinner units (predominantly GC, URC, PGC and LRC), the stratigraphic sequence of (from top to bottom) LC + JUIF + GC + URC + PGC + LRC + LRGC can be recognized with similar thickness data for all intercepts in the same unit. That stratigraphic sequencing is not arbitrary since it is supported by a mineral signature particular to each unit i.e. medium Fe + med magnetite in LC, high Fe + high magnetite in JUIF, low Fe + low magnetite in GC, high Fe + medium-high magnetite in URC, medium Fe + mediumhigh magnetite in PGC, high Fe + high magnetite in LRC, high Fe + medium-high magnetite in LGRC. Given the well-documented geological interpretation of the area by NML and; the relative low variability of the Fe_Head data, as well as the overall continuity of the mineralization between sections and the 70 holes on the 1000m x 500m grid, SGS has elected to classify the mineralized material recognized by those holes in the indicated category. Hence the mineral inventory made of blocks within the 1000m x 500m cells of influence of holes is classified as indicated. The balance of the mineral inventory is classified in the inferred category. Resource Estimation- Final estimates Resources are that part of the mineral inventory with a reasonable prospect of economic extraction which means both a cut-off grade and an optimized pit shell i.e. blocks with estimated values above

11 Resource Estimation of the Sheps and Perault Taconite Properties Page 6 a cut-off and within the limits of an optimized pit shell. Traditionally, the cut-off used to report resources in the taconite deposits of the Labrador Trough is a minimum 18% weight recovery of the magnetic concentrate from Davis Tube test on material ground to 325 mesh. According to BBA (2009) the unit cost and concentrate values used in the PFS of KéMag suggest a lower marginal cutoff. In that study, a pit optimization is run with a C$4.03/t ROM total processing + G&A cost and a concentrate value of C$49.92/t CC (unit mining cost is C$1.75/t ROM) hence a marginal weight recovery cut-off of : 4.03/49.92 = 8%. A Whittle optimized pit shell based on the mineral inventory described in previous sections (both indicated and inferred blocks), the above NML cost and value parameters and maximum pit slope angles of 50. The optimized pit shell includes most of the mineral inventory and all the blocks with an estimated weight recovery above 15% recovery cut-off. As a result, the proposed resources for Sheps-Perault are made of all the blocks of the mineral inventory above 15% minimum recovery cut-off (not restrained by any pit shell). These mineral resources are listed below by unit, total and total after merging unit fractions in blocks. As expected from the statistics of sample data in each unit, there are no resources above 15% recovery in the GC. Further work is required to potentially upgrade these mineral resources to mineral reserves, and it has not yet been demonstrated that these mineral resources have economic viability. Estimated Resources Sheps & Perault COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % 15 Indicated 1,208 4, Inferred 301 1, Estimated Resources Sheps COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % 15 Indicated 596 2, Inferred Estimated Resources Perault COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % 15 Indicated 613 2, Inferred Dated February 7 th, 2013 Estimated resources of the Sheps Lake & Perault Lake Properties Conclusion and Recommendation New Millennium Iron Corp. (NML) holds the 12,597 ha Sheps Lake and Perault Lake properties about 130 km to the northwest of Schefferville in Labrador (NL) close to the northern Québec (QC) provincial boundary and along the so-called Millennium Iron Range (MIR) that comprises their LabMag and KéMag deposits to the southeast and extends to the Lac Otelniuk deposit of Adriana Resources to the northwest. On these properties, NML has drilled a 35 km stretch of the MIR corresponding to a strong magnetic anomaly recognized by a 2010 airborne survey by Fugro and bounded at its southeast and northwest extremities by interpreted vertical fault zones. 73 vertical holes totaling 5,970 m have been drilled on 8 SW-NE sections 2 km apart, mostly having one hole per section; and on 19 SW-NE sections 1 km apart and with a spacing of about 0.5 km between holes on the same section (average of 3 per section). The main unit is the top Lean Chert (LC)

12 Resource Estimation of the Sheps and Perault Taconite Properties Page 7 followed from top to bottom by the Jasper Upper Iron Formation (JUIF), the Green Chert (GC) marker horizon, the Upper Red Chert (URC) unit, the Pinky Green Cert (PGC) unit, the Lower Red Green Chert (LRGC) unit and finally the Lower Red Green Chert (LRGC) unit for a total average thickness of about 70 m. The thickness of individual units does not vary much from hole to hole but is lost during several sections. However, the geological continuity over kilometric distances is well demonstrated. The unit package dips gently by 10 to the northeast on all sections. A statistical analysis of sample returns by lithological units shows that each unit has a specific geochemical + mineralogical signature i.e. medium Fe + med magnetite in LC, high Fe + high magnetite in JUIF, low Fe + low magnetite in GC, high Fe + medium-high magnetite in URC, medium Fe + medium-high magnetite in PGC, high Fe + high magnetite in LRC, high Fe + medium-high magnetite in LGRC. Despite the long spacing between holes, the grade continuity is demonstrated by the low variability of the % Fe Head data as well as the geological information and interpretations from NML. There is however, an observed variability in the recovery data that can be explained by the difference between the Sheps (higher recovery) and the Perault (Lower recovery) property areas. Additional drilling and/or geological data gathering and interpretation would be necessary in order to ascertain this difference. The SGS independent QP collected 10 independent check samples during his site visit that were then sent to SGS Lakefield. Compared to the MRC data, the SGS Lakefield data for %Fe of concentrate appears to be significantly more (by about 96% Fe) and for % SiO 2 significantly less (by about 80% SiO 2 ). This discrepancy might be linked to the way these elements have been analyzed (titration or wet chemistry at MRC and XRF at Lakefield). Although the standard operating procedures from MRC and SGS-Lakefield are similar, some differences might have been unintentionally created from grinding (size reduction) or from calibration of the magnetic separator (Davis Tube). A detailed investigation of these differences is recommended. SGS considers that the number of samples checked (10) is relatively small and conclusions should be interpreted accordingly. SGS recommends increasing the number of check samples in future verifications when more holes are available. Unit volumes in blocks are converted into tonnages using a calculated density for every block according to %Fe of Head. The calculated densities are derived from a linear regression formula based on % Fe of Head but restricted to each unit. The measurements made on the similar deposit, which is under feasibility, vary per unit with a average of 3.29 t/m 3 in LC, 3.43 t/m 3 in the JUIF, 3.15 t/m 3 in the GC, 3.53 t/m 3 in URC, 3.43 t/m 3 in the PGC, 3.40 t/m 3 in the LRC and finally 3.33 t/m 3 in LRGC. The calculated densities show a strong to moderate correlation with the % Fe of head but are not derived directly from the Sheps and Perault main deposits area. Estimated resources are made of blocks with a combined interpolated weight recovery of all units in the block above the traditional cut-off of 18% weight recovery. Blocks above that cut-off are limited to the LC and URC units. That traditional cut-off is the one used to report resources of NML s LabMag and KéMag deposits to the southwest as well as Adriana s Lac Otelniuk deposit to the northeast. Given the most recent figures for concentrate values and unit processing costs, the 18% minimum weight recovery was reduced to 15%. A Whittle optimized pit shell based on those figures as well as a 50 maximum slope includes most of the indicated and inferred blocks of the mineral inventory above the 15% weight recovery cut-off, thus demonstrating their reasonable prospect of economic extraction. SGS Geostat offers the following recommendations for further evaluation of the Sheps Lake and Perault Lake Properties:

13 Resource Estimation of the Sheps and Perault Taconite Properties Page 8 Measured densities are currently restricted to the LabMag and KéMag properties based on the feasibility in SGS recommends carrying density measurements on the Sheps and Perault properties with standard water immersion of core fragments and pycnometer on pulps. If a higherconfidence relationship can be made with more density measurements on individual samples in each unit, it could replace the ones used in the resource model to (1) better combine estimates of different litho unit fractions in the same block (2) have tonnage and to some extent weight recovery estimates above cut-off that would reflect a slight expected increase of density with the weight recovery cut-off since we have a mild positive correlation of weight recovery and %Fe of head. Practically, we suggest submitting about 250 pulps rejects (45 for LC, 45 for LRGC and at least 30 for the other units using a large range of %Fe of head in each unit) to a pycnometer measurement. Those results would then be used to build regression models of density with %Fe of head in each unit and then to review the current resource block model with block densities derived from those regression models. The estimated and conceptual cost of this operation is about around C$15,000. An economic analysis (PEA or PFS) should be conducted with the currently estimated indicated and inferred resources. This study would help determine with higher confidence the economic factors such as product value and unit mining/processing costs and lead to more robust affirmation that a realistic cut-off of 15% minimum weight recovery is adequate to report resources. The estimated and conceptual cost of the economic study is ranging from C$150,000 to C$275,000. Although the current drill hole data demonstrate a sufficient continuity of both geology and grade for a classification of all the material recognized by holes on the 1x0.5 km grid in the indicated category, we recommend another drilling programs before starting a preliminary feasibility study (PFS) with the currently estimated indicated resources. That program would have a magnitude similar to the 2011 drilling program i.e. with about 35 NQ holes totaling about 3,500 m. The majority of the new holes (about 20) would be drilled on the 5 (2 km spaced) sections to the NW where only one DDH is present on average as well as the intermediate sections between them. To the SE, The drilling campaign should focus also on the drilling (about 5) of the intermediate 500 m sections with the same spacing of 500 m between holes on the same section depending on the grade and continuity. The balance of about 10 holes would be on the SE sections where no drilling was done. The purpose of that additional drilling is not to put the estimated resources in the measured category but to validate the apparent geological and grade continuity of the first 40 holes. With about 100% more data, geological discontinuities (barren dikes, faults, etc) may show up and the spacial distribution may change significantly. Also with a 500m x 500 m drilling grid, the indicated resources would have a drilling density similar to indicated resources of Lac Otelnuk, LabMag and KéMag deposits to the north, all of which are in a similar geological environment. Check samples were sent to the SGS Lakefield lab for preparation and Davis Tube testing withthe same standard operation procedure ( SOP ) to that of the Midland Research Center lab. The Four variables were reported and compared. According to a series of tests, a bias was observed on all of the four variables. These apparent differences between MRC and SGS Lakefield for the quality of concentrate and Fe Head are somewhat similar to observations on the NML s Lac Ritchie property ( NML Technical Report, 2012) and may be linked to the way that quality is measured at the labs (titration and thermo-gravimetry at MRC and XRF for SGS Lakefield). Although the MRC and SGS-Lakefield SOPs are similar, some differences might have been unintentionally created from grinding (size reduction) or from calibration of the magnetic separator (Davis Tube). A detailed investigation of these differences is recommended. At his stage the author is unable to estimate any costs for this study.

14 Resource Estimation of the Sheps and Perault Taconite Properties Page 9 2 Introduction 2.1 General This technical report was prepared by SGS for NML to support the first disclosure of mineral resources completed for the Sheps Lake and Perault Lake properties. The report describes the basis and methodology used for modeling and estimating of the Sheps Lake and Perault Lake resources. The report presents a summary of the history, geology, sample preparation and analysis, data verification and metallurgical work completed on the Sheps Lake and Perault Lake properties ( Properties or Property ). The report also provides recommendations for future work. 2.2 Terms of Reference This report was prepared by Maxime Dupéré, P.Géo. Mr. Dupéré is responsible for the site visit, independent sampling verification, resource estimation and all sections of this technical report. Mr. Dupéré acknowledges the help of Mr. Michel Dagbert, Eng. and senior geostatistician for SGS Geostat for his helpful advice on the mineral resource estimates in this technical report. This technical report was prepared according to the guidelines set under Form F1 Technical Report of National Instrument Standards and Disclosure for Mineral Projects. The certificate of qualification for the Qualified Person responsible for this technical report have been supplied to NML as a separate document and can also be found at the very end of the report. Mr. Dupéré visited the Property between August 20 th to 22 nd 2012, for a review of exploration methodology, sampling procedures and to conduct an independent check sampling of selected mineralized drill intervals. 2.3 Source of Information Information in this report is based on critical review of the documents, information and maps provided by personnel of NML, in particular T. (BK) Balakrishnan, P.Geo., Chief Geologist and Michael Spleit, Eng, Mining Engineer. Drilling data was provided by NML and validated against information obtained during the field visit and certificates from the analytical laboratories. 2.4 Currency, Units, abbreviations and Definitions All measurements in this report are presented in the Système International d Unités (SI) metric units, including metric tonnes (tonnes) or grams (g) for weight, meters (m) or kilometers (km) for distance, hectare (ha) for area, and cubic metres (m 3 ) for volume. All currency amounts are Canadian Dollars (C$) unless otherwise stated. Abbreviations used in this report are listed in Table 2.1.

15 Resource Estimation of the Sheps and Perault Taconite Properties Page 10 Table 2.1: List of Abbreviations Tonnes or t kg g km m Metric tonnes Kilograms Grams Kilometers Meters µm Micrometers ha Hectares m³ Cubic meters % Percentage $ Dollars (CAD unless otherwise specified) Degree C Degree Celsius ppm BQ NQ SG UTM Parts per million Drill core size (3.65 cm in diameter) Drill core size (4.76 cm in diameter) Specific Gravity Universal Transverse Mercator

16 Resource Estimation of the Sheps and Perault Taconite Properties Page 11 3 Reliance on Other Experts The author of this Technical Report is not qualified to comment on issues related to legal agreements, royalties, permitting, and environmental matters. The author has relied upon the representations and documentations supplied by the Company s management. The author assumes that the documents, reports and other data listed are substantially accurate and complete in all material aspects. The author has reviewed the mining titles (claims), their status, the legal agreement and technical data supplied by the Company, and any public sources of relevant technical information.

17 Resource Estimation of the Sheps and Perault Taconite Properties Page 12 4 Property Description and Location 4.1 Location The Sheps Lake and Perault Lake Properties are located approximately 20 km and 17 km southwest of Schefferville, QC, 220 km north of Labrador City, NL and 520 km north of Sept-Îles, QC. The Properties are situated in unorganized territory, straddling NTS map sheets 23J10, 23J11, 23J14 and centered at W, N. There has been no mining activity on the Properties and as such there are no mine workings, tailings impoundment areas, waste piles or other infrastructure on or near the Property. The Property location is shown on Figure 4.1.

18 Resource Estimation of the Sheps and Perault Taconite Properties Page 13 Figure 4.1: Location map of the Sheps Lake & Perault Lake properties

19 Resource Estimation of the Sheps and Perault Taconite Properties Page Property Description and Ownership The two Properties consist of a total of 120 contiguous mineral rights licences totalling 562 claims and km². The area of each map-staked claim is 25 hectares or less. The claims have not been marked out on the ground or legally surveyed, nor is there any requirement to do so. NML detains 100% of the Perault property. NML detains 80% of interests and Naskapi Nation of Kawawachikamach (NNK) 20% of the Sheps property. There are no royalties on either property. In Newfoundland & Labrador, claims are valid for five year periods and convey only mining rights, no surface rights. To maintain claims in good standing, claims must be renewed prior to their expiry date. Renewal requires the filing with the ministry of acceptable work expenditures in the form of a technical report and the payment of a fee. The claims may be held for a maximum of twenty years. A minimum assessment work is required between C$200 and C$1,200 per claim per year depending on the year and term of the licence. Excess assessment work completed in any one year is carried forward for a maximum of nine years and it is automatically credited to the licence. Some of the NML claims have been renewed once and others are in the process of being renewed. NML does not hold surface rights to the Properties nor are they required at this stage of the Property s development. The Properties location map and holdings are respectively shown on Figure 4.1 and in Table 4.1. Additional information is available in the Adjacent Properties section.

20 Resource Estimation of the Sheps and Perault Taconite Properties Page 15 Table 4.1: Summary of NML mineral claims holding at Sheps Lake & Perault Lake NTS Map sheets Licence No. of Claims Area Hectare Stake date Area Ownership 23J10, 23J M /15/2004 Perault Lake NML (100%) 23J M /09/2007 Perault Lake NML (100%) 23J10, 23J M /24/2007 Perault Lake NML (100%) 23J M /18/2008 Perault Lake NML (100%) 23J10, 23J M /15/2011 Perault Lake NML (100%) 23J10, 23J M /15/2011 Perault Lake NML (100%) 23J M /15/2011 Perault Lake NML (100%) 23J M /15/2011 Perault Lake NML (100%) 23J M /15/2011 Perault Lake NML (100%) 23J14, 23J M /15/2004 Sheps Lake NML (80%), NNK(20%) 23J M /01/2006 Sheps Lake NML (80%), NNK(20%) 23J M /10/2007 Sheps Lake NML (80%), NNK(20%) 23J14, 23J M /10/2007 Sheps Lake NML (80%), NNK(20%) 23J M /06/2008 Sheps Lake NML (80%), NNK(20%) 23J M /06/2008 Sheps Lake NML (80%), NNK(20%) 23J M /27/2005 Sheps Lake NML (80%), NNK(20%) 23J14, 23J M /15/2011 Sheps Lake NML (80%), NNK(20%) 23J M /15/2011 Sheps Lake NML (80%), NNK(20%) 23J14, 23J M /15/2011 Sheps Lake NML (80%), NNK(20%) 23J M /28/2012 Sheps Lake NML (80%), NNK(20%) Total Claims 562 Total Area Permits For its diamond drilling program, NML required work permits from Newfoundland & Labrador Government, Approval E120023, dated April 12, It was valid till December 31 st, 2012.

21 Resource Estimation of the Sheps and Perault Taconite Properties Page 16 Figure 4.2: Sheps Lake & Perault Lake Properties & Others Owners

22 Resource Estimation of the Sheps and Perault Taconite Properties Page 17 5 Accessibility, Climate, Local Resources, Infrastructure and Physiography 5.1 Accessibility Access to exploration drilling area was by truck and snowmobile. A few lakes around to the Properties are accessible from Schefferville via chartered fixed-wing float aircraft and such aircraft was used to support NML s 2011 and 2012 drilling programs along with a helicopter for drill crew moves. During 2012, the drilling program equipment was transported on site by helicopter and with a dozer. There is daily scheduled air service to Sept-Îles and from there to Québec City, Montréal and beyond. There is twice-a-week round-trip train service for passengers and freight between Schefferville and Sept-Îles. 5.2 Climate The Properties area has a sub-arctic climate with temperatures which average 12 C in July and 25 C in December. The average annual temperature is -6 C. Average annual rainfall is approximately 410 mm and snowfall 440 cm. 5.3 Local Resources and Infrastructure The Properties have no inhabitants. The nearest Hydro-Québec transmission lines are in Schefferville, where local needs are served by hydroelectric power from the Menihek Lake power plant located nearby in Labrador. There is a more than adequate supply of water available for exploration and mining purposes; however, there is no harvestable timber on each Property. There is ample room available on each Property for the establishment of mining and processing operations, waste piles and a tailings management area. Schefferville, Québec is the closest population centre and has a population of approximately 300. The Matimekush (Innu) Reserve is contiguous with the village and is in effect a part of it. The total Schefferville area population is approximately 1,500 including that of the Kawawachikamach (Naskapi) Reserve, which is a few km east of Schefferville by road. There is a very small, unskilled labor pool in Schefferville. Extensive training would be required for any mining operation and the bulk of the workforce would have to come from the south. Schefferville was built in the early 1950s to serve as the residential and service centre for the Iron Ore Company of Canada ("IOCC") iron mining operations and is the northern terminus of the Quebec North Shore & Labrador Railway ("QNS&L"). There are several stores, a hotel, a "Bed and Breakfast"-type inn, a restaurant and some services available. There are primary and secondary schools and a health clinic. There are dwellings available for rent, a seasonal charter float-plane service and there are daily scheduled flights to Sept-Îles in a small commercial aircraft. The village is served by reliable hydroelectricity and there is twice-a-week rail service to and from Sept-Îles. The km journey takes approximately 15 hours one-way and delays are frequent as trains hauling

23 Resource Estimation of the Sheps and Perault Taconite Properties Page 18 iron ore concentrate and pellets from Wabush to Sept-Îles have priority on the line. Ross Bay Junction (the Wabush corner) is 228 km south by rail from Schefferville and Sept-Îles is a further km. The rail bed from Ross Bay Junction to Schefferville has deteriorated since 1982 when IOCC closed its Schefferville operation and heavy-duty rail was replaced by lighter-gauge rail. A consortium of First Nations groups purchased the Ross Bay to Schefferville portion of the line and operating under the name of Tshiuetin Railways. It is expected that there will be Federal Government assistance in upgrading the track for heavy duty transportation of NML/TSMC DSO ore and Labrador Iron Mine which is now shipped in small volumes. 5.4 Physiography Topography is flat to gently rolling from west to east and is rough, with the occasional more precipitous area. For Sheps Lake, the taconite is well exposed as a long linear belt 1.5 to 2 km. The properties well drained, has a few swampy areas and is covered by sparse northern boreal forest consisting of stunted spruce, alders and willows. The properties are dominated by tundra with rare stunted black spruce.

24 Resource Estimation of the Sheps and Perault Taconite Properties Page 19 6 History The first geological work known was done in 1952 by Dr. Perrault who was employed by Iron Ore Company of Canada (IOC). He mapped this property by the pace and compass method, no assay samples were taken at that time. The description of formation was: Menihek, Sokoman, Ruth and Wishart. Dr. Perrault was assumed that the iron formation is cut by one cross fault between Sheps and Perault. In November 1960, Robert A. Martin did the iron economical evaluation of Perault Lake area; 11 lines were surveyed by Aravella magnetometer. Outcrops were described and updated from previous basic regional work. Representative samples were taken for assaying but it was impossible to determine if enrichment and leaching continued at depth. The samples had no economic importance but were necessary in order to complete the evaluation. There has been no mining activity or exploration program before Perault Lake exploration by drilling in In the summer of 2011, NML, 2 holes were drilled for a total of 3,500 m but no exploration for Sheps Lake before 2012.

25 Resource Estimation of the Sheps and Perault Taconite Properties Page 20 7 Geological Setting and Mineralization 7.1 Geological Setting The following information is based on observations by Thiagarajan Balakrishnan P.Geo., NML s Chief Geologist involved in the assessment report of exploration for the magnetic Taconite in Labrador (NL). Mr. Balakrishnan is a member of the PEGNL (Professional Engineers and Geoscientists Newfoundland and Labrador) Regional Geology The Property is situated in the Churchill Province, of the Labrador Trough ("Trough") adjacent to the Archean basement gneiss (Figure 7.1). The Trough, otherwise known as the Labrador-Québec Fold Belt, extends for more than 1,000 km along the eastern margin of the Superior Craton from Ungava Bay to Lake Pletpi, Québec. The belt is about 100 km wide in its central part and narrows considerably to the north and south. The Trough comprises a sequence of Proterozoic sedimentary rocks, including iron formation, volcanic rocks and mafic intrusions. The southern part of the Trough is crossed by the Grenville Front representing a metamorphic fold-thrust belt formed during the 1,000 Ma Grenvillian Orogeny. Trough rocks in the Grenville Province are highly metamorphosed and complexly folded. Iron deposits in the Grenville part of the Trough; include Lac Jeannine, Fire Lake, Mont-Wright, Mont-Reed, and Bloom Lake in the Manicouagan-Fermont area and the Luce, Humphrey and Scully deposits in the Wabush-Labrador City area shown on Figure 7.1. The high-grade metamorphism of the Grenville Province is responsible for re-crystallization of both iron oxides and silica in primary iron formation, producing coarse-grained sugary quartz, magnetite, and specular hematite schists (meta-taconites) that are amenable for coarse grinding and concentration by gravity methods.

26 Resource Estimation of the Sheps and Perault Taconite Properties Page 21 Figure 7.1: Geological map of the Labrador Through

27 Resource Estimation of the Sheps and Perault Taconite Properties Page 22 NML s Sheps Lake and Perault Lake Properties are located north of the Grenville Front in the Churchill Province where the Trough rocks have been only subject to greenschist or sub-greenschist grade metamorphism. The principal iron formation unit is known as the Sokoman Formation. The Sokoman Formation, member of the Ferriman Sub-Group, is overlain by the Menihek Sub-Group (mudstone and shales) and underlain by the Wishart Formation (quartzite), the Denault Formation (dolomite) and the Attikamagen Formation (shale). The regional stratigraphic column is shown on Table 7.1. Iron formations in this part of the Trough are taconites, which are weakly metamorphosed. Magnetic taconite iron deposits in the Trough include NML s KéMag and LabMag deposits (Howells River Deposit) and the December Lake deposit. The Direct Shipping Ore deposits ( DSO ) occurring near Schefferville were derived from highly folded and faulted taconite iron formations which were leached of silica and other gangue minerals by percolating meteoric waters and by secondary enrichment. Table 7.1: Regional Stratigraphic Column Eon Age Super Group Sub-Group Formation Unit Menihek Shale LC Sokoman GC URC Proterozoic Aphebian Kaniapiscau Knob Lake Ferriman LRGC JSP Ruth RS BC Wishart Quartzite Attikamagen Swampy Bay Pistolet Seward Archean After Dimroth, 1978 Churchill Province

28 Resource Estimation of the Sheps and Perault Taconite Properties Page Property Geology General The Properties are located on the western edge of the Trough. Archean gneisses form the basement and dip gently east. The basement gneisses are unconformably overlain by the gently northeast dipping sedimentary rocks defined as the Kaniapiscau Supergroup, which includes the Ferriman Sub-Group and iron-bearing formations belonging to the Sokoman Formation. Towards the western edge of the Trough, the older, lower units of the sequence are successively exposed as the upper younger units have been removed by erosion. To the northeast, the Ferriman rocks are overlain by the Menihek shales (Figure 7.2). Within the Sheps Lake and Perault Lake properties the structure is very simple; the iron formation is generally northwest-southeast striking and dipping 5 to 10 to northeast. Metamorphism within the Property appears to be of low to moderate grade. Changes in grain size mineralogy and rock texture related to metamorphism are not visually identifiable. Within oxide iron formation units, the most notable feature in mineral composition is the rather abrupt changes from dominantly magnetite (LC) to dominantly hematite (URC, JSP), and corresponding change of the silica from chert over to jasper. These oxidation potential variations and changes in iron grade define the member lithology units. The occurrence of iron carbonate minerals, principally siderite, ferro-dolomite and ankerite are widespread but are more abundant in the upper (LC) and lower (LRGC) units. Iron silicates, minnesotite and stilpnomelane occur in LC, GC and LRGC units. These features all appear to be related to primary deposition.

29 Resource Estimation of the Sheps and Perault Taconite Properties Page 24 Figure 7.2: Sheps Lake and Perault Lake geology

30 Resource Estimation of the Sheps and Perault Taconite Properties Page Lithology The taconite iron formation in the Sheps Lake and Perault Lake area is part of the Sokoman Formation occurring at the western margin of the trough. The taconite is well exposed as a long linear belt 1.5 to 2.0 km wide for a total length of approximately 6 km in a northwest-southeast trend which constitutes the AM anomalous zone. Among the members of the Knob Lake Group, the Attikamagen Formation and the Denault Formation are not exposed and absent all along the 210 km long magnetic iron ore (taconite) belt called the Millennium Iron Range (MIR) located near Schefferville, Quebec. The Ruth Formation followed by Sokoman Formation overlie the Wishart Formation, which is a fine to medium grained sand stone sometimes arkosic containing feldspar grains. The Wishart Formation overlies unconformably the basement Archean gneisses. The following stratigraphic sequence is established in the properties investigated (Table 7.2). Table 7.2: Detailed taconite stratigraphic column for the Property area Unit Menihek Formation (Shale) (MS) Sokoman Formation UIF Member Estimated Average True Thickness and Range (m) > 79.2 Thrust Fault Description Dark grey to black shale with minor interbedded greywacke and carbonate lithofacies, carbonaceous pyretic shale. Lean Chert Sub-member (LC) Jasper Upper Iron Formation (JUIF) Magnetite-Carbonate Facies Green Chert (GC) Magnetite-Carbonate Facies 25.0 ( ) 26.2 ( ) 3.8 ( ) Greenish, green to grey-green and pink-grey magnetite-chert iron formation with local zones of laminated to shaley bedded (siderite-magnetite) chert iron formation. This unit contains stromatolite bearing purple-red and green chert band with magnetite less than 3m thick. Stilpnomelane bearing magnetite-rich shales occur both above and below the stromatolitic band. Layered to laminated, magnetite-chert iron formation. Red-grey-pink in colour, red chert and oolites. Silicate-rich, green chert unit, laterally continuous and an excellent marker horizon.

31 Resource Estimation of the Sheps and Perault Taconite Properties Page 26 MIF Member Unit Upper Red Cherty (URC) Hematite-Carbonate Facies Pink-Grey Cherty (PGC) Magnetite-Carbonate Facies Estimated Average True Thickness and Range (m) 8.1 ( ) 12.6 ( ) Description Massive to layered, jasper-magnetite-chert iron formation. Red-grey to reddish purple. Disseminated magnetite-chert iron formation. Grey to pink-grey to green-grey. Lower Red Cherty (LRC) Hematite-Carbonate Facies LIF Member Lower Red Green Cherty (LRGC) Magnetite-Carbonate Facies 8.6 (0-18.6) 21.2 (0-46.0) Layered magnetite-chert iron formation. Red-grey to reddish purple. Lower contact transitional. Layered silicate-magnetite-carbonate, magnetite-chert iron formation. Pink to reddish-grey to green-grey. More silicates in lower part, more oxides in upper part. Lower contact transitional with LIF. Lower Iron Formation (LIF) Silicate Facies 8.2 ( ) Massive to layered green to grey-green silicate-carbonate-magnetite-chert iron formation. Ruth Formation (RF) 5.2 ( ) Thin bedded to laminated chert-siderite, with thin bands of shale. Note Zajac (1974) argues the term Ruth Formation should be abandoned because it is for most Sulphide Facies part equivalent to LIF. Black chert 1.4 m ( m). Wishart Formation 17.7 ( ) Quartzites and /or re-crystallized cherts. UNCONFORMITY Ashuanipi Complex Archean Granites and Granodioritic gneiss and mafic intrusives. Paleosol on contact between Proterozoic Assemblage and Archean basement.

32 Resource Estimation of the Sheps and Perault Taconite Properties Page Structure The Taconite formation at Sheps Lake and Perault Lake area is gently dipping at 5 to 10 to northeast. The Wishart, Ruth and Sokoman Formations are un-deformed and strike at a northwestsoutheast direction. However, two cross faults dissects the iron formation in approximately northeast-southwest directions. These faults are vertical dip-slip faults extending from the basement gneisses in the west and well into the iron formation and Menihek formation in the east. These faults form narrow, deep erosional valleys. Because of significant uplift and subsequent erosion, displacement in the dip direction is considerable and easily noticeable. Due to heavy fracturing close to either side of the faults, the iron formation is altered by percolating meteoric waters and the oxidation of magnetite to hematite (martite) is significant. Most of the carbonates and silicates are altered to limonite and goethite. Because of the altered nature of the iron formation the core recovery is very poor in holes drilled close to the fault zone. However, in spite of wide spread weathering no significant leaching of silica (SiO2) had occurred.

33 Resource Estimation of the Sheps and Perault Taconite Properties Page Mineralization The taconite iron formation consists mostly of varicolored recrystallized cherts, jasper with bands and disseminations of magnetite. Hematite in its original form is also present in most of the units. Martite, an oxidized form of magnetite is present where the iron formation is weathered and highly altered. The gangue minerals present in addition to chert are iron silicates minnesotaite and stilpnomelane, iron carbonates siderite, ankerite and ferrodolomite. The carbonate, calcite occurs mostly as fracture fillings especially in Menihek slate. Units LC, PGC, LRC and LRGC show higher concentrations of magnetite, while URC and JUIF tend to contain higher amounts of hematite compared to all other units. Iron carbonates occur in LC, GC and LIF units. Iron silicates are common in LC, GC and LIF units. The green color of GC is mainly due to the presence of minnesotaite. Minor manganese carbonates rhodochrosite and kutnohorite are reported also to be present.

34 Resource Estimation of the Sheps and Perault Taconite Properties Page 29 8 Exploration During 2010 New Millennium Iron Corp ( NML ) carried out Airborne High Resolution Magnetic Survey ( AM ) over the Millennium Iron Range ( MIR ) to identify additional, potential magnetic taconite deposits similar to the proven LabMag deposit in the Howells River Area and the KéMag deposit in Québec. The survey covered the taconite belt over a length of 150 km starting south of Perault Lake, to north of Lac Ritchie in Quebec. Fugro Airborne Surveys Corp. of Ottawa, Ontario ( Fugro ) carried out the airborne survey between October 15 th and November 11 th, 2010 and submitted the Logistics and Processing Report, Airborne High Resolution Magnetic Survey. Mr. Jean M. Hubert, Geophysical Consultant for NML did the interpretation of the airborne survey data provided by Fugro and submitted the Report of Analysis of An Airborne Survey on the LabMag Property. Several prominent magnetic anomalies were outlined and recommended for detailed exploration work. Two such anomalies cover the Sheps and Perault Lake areas which occur as the southern extension of the LabMag taconite belt (Figure 8.1).

35 Resource Estimation of the Sheps and Perault Taconite Properties Page 30 Figure 8.1: Taconite targets on the residual magnetic grid in the Sheps Lake and Perault Lake area

36 Resource Estimation of the Sheps and Perault Taconite Properties Page 31 9 Drilling 9.1 Sheps Lake property Based on the presence of strong magnetic anomalies of 2010, a detailed exploration drilling program was undertaken in the Sheps Lake and Perault Lake AM anomalies area during the spring The drilling contractor is Georges Downing Estate Drilling Ltd. of Grenville-Sur-la-Rouge, QC. The area covered by major part of the drilling is approximately 5 km long and 1 to 2 km wide and encompasses licences M, M, M, M and M (Figure 9.1). The program started on April 20 with 3 drills and concluded on May 30, During that period 23 holes were drilled totaling 1,920.1 metres in length. The drilling was carried out on a grid of lines spaced 1 km apart with 3 to 4 holes spaced 300 to 500 m on each section line. Prior to drilling all the holes were spotted on the ground using handheld GPS unit. Several selected samples of crude ore and Davis Tube concentrate samples were sent to SGS Lakefield, Ontario, for elemental analysis by XRF method. At the conclusion of drilling all the drilled holes were re-surveyed by N. E. Parrott Surveys Ltd. of Happy Valley-Goose Bay, NL.

37 Resource Estimation of the Sheps and Perault Taconite Properties Page 32 Figure 9.1: Drilling program and sections lines of the Sheps Lake area

38 Resource Estimation of the Sheps and Perault Taconite Properties Page Perault Lake property The area covered by drilling is approximately 18 km long and 1 to 2 km wide and encompasses licences M and M (Figure 9.2). The program started with one drill and subsequently joined by 2 additional drills provided by the contractor Downing Estate Drilling Ltd. It was decided to drill initially on lines spaced 2 km apart starting from section line 130 in the south with 2 holes spaced 300 to 400m apart on each line. This pattern was followed till the drills reached section line 200. Between sections 200 and 310, drilling was carried out on lines spaced 1 kilometre apart 3 to 4 holes spaced 200 to 300 metres along section lines. For the 2012 program, logging, sampling and related work were completed by seniors geologists Henry Simpson, Bocar Diagana, Junior geologist Alex Howe and student geologists Julie Andrews, William (Billy) Clarke and Jonathon Remedios, Chris Baker and Patrice Mathieu of Memorial University in St. John s, NL. Logging was done using an MS Excel template and then imported in to a MS Access database. Descriptive core logging used the lithology codes developed by IOCC and used by NML for LabMag and KéMag properties. Core logging included Rock Quality Index ("RQD"), magnetic susceptibility measurements in 0.15 m intervals down the core and core photography. Sample intervals are marked on the core by the logging geologists using china markers or lumber crayons and then recorded in 3-part sample books. The entire iron-rich section of the drill core was sampled leaving no gaps. Sample lengths were based on geological criteria and sample lengths averaged approximately 5.7 m; minimum 0.7 m, maximum 9.9 m. These sample intervals are similar to what was done at KéMag and LabMag deposits. One portion of the 3-part sample tickets are stapled into the core trays at the beginning of each sample interval. Core splitting was done using three Hydrasplit hydraulic splitters. Split core samples were placed into nylon rice bags with the second portion of the 3-part sample tags and tied securely with plastic ties along with a yellow tag with the sample No. marked. The collected samples were then packed into 45 gallon steel drums, labelled with a sequential drum number and the address of the Laboratory. Samples were sent in batches from Schefferville by train to Sept-Iles, where the sample drums were picked up for transportation by truck to MRC laboratory. The save cores are temporarily stored at Schefferville and eventually will be sent to Labrador City for permanent storage.

39 Resource Estimation of the Sheps and Perault Taconite Properties Page 34 Figure 9.2: Drilling program, section lines and geology of the Perault Lake area

40 Resource Estimation of the Sheps and Perault Taconite Properties Page Sample Preparation, Analyses and Security All the split core samples were sent to the Midland Research Center (MRC) in Nashwauk, Minnesota for chemical and Davis Tube (DT) analysis. A total of 326 core samples including 17 duplicate check samples were collected on the Sheps Lake property and a total of 618 core samples including 32 duplicate check samples were collected on the Perault Lake property. The following test work and sample analyses were completed for all samples by MRC: Head assay for total iron (%FeH) Determination of magnetic weight recovery (%WRec) on -325 mesh concentrate Determination of total iron (%FeC) and silica (%SiO 2 C) in concentrate MRC sample preparation and analysis flowsheet consisted of the following steps: Individual core samples crushed to 3/8 with a 4 x6 jaw crusher Split of 1500g for test work Save the balance Roll crush 1500g to 100% -10 mesh Split 50g for DT test and head sample analysis Save the balance Stage grind 50 g to -325 mesh as per MRC procedure (Hanna Procedure) DT weight recovery test on g sample as per MRC procedure (Hanna procedure) Analyze DT concentrate for total %Fe and SiO 2 (non-mercury titrimetric method for total iron and SiO 2 determination using hydrofluoric acid) The security measures to protect the sample integrity are adequate and consist in identifying of sample bags with drill hole name, from-to and sample number, referencing sample locations in core boxes and direct shipment of sample bags containing half core pieces to the MRC laboratory. No sample preparation was done on site NML QA/QC Duplicate samples at MRC Over the course of the drilling and sampling program, NML has sent 49 duplicate samples to the Midland Research Center lab. Those samples are actually the second half of the core. At MRC, those samples were subjected to the same Davis Tube testing as the original half core for the same intervals. Duplicate samples are from 47 of the 73 holes. Multiple samples in the same hole are in holes 12PL1018D (2). Their length varies from 1.9 to 9 metres with an average of 5.4 metres. Most of duplicate samples are in the LRGC unit (15) followed by the LC (9), the JUIF, the PGC and the URC units (5 in each), the LRC unit (4), the GC unit (1) and the LIF units (3 in each). Given that

41 Resource Estimation of the Sheps and Perault Taconite Properties Page 36 reported resources are mostly in the LRGC and LC units with some in the URC unit (see chapter 14), future duplicate samples should concentrate on those units. Results for duplicates include the same variables as the original samples i.e. the %Fe of head, the %Weight recovery, the %Fe of concentrate and the %SiO 2 of concentrate. Correlation plots of duplicated values and original ones are on Figure As a general rule; the duplicated values reproduce reasonably well the original data. Sign tests and T-tests of paired dated do not show any significant bias for all four parameters except for the SiO 2 Concentrate where a bias was observed from the results of the Student T Test. (Table 10.1). This bias may be due to the presence of an outlier sample data (#7919). By removing it from the sample series the presence of the bias was lifted, SGS recommends investigating and commenting on the different data that are or should be removed from the quality control list. Table 10.1 Statistics of NML duplicated data Variable %Fe_Head %WtRec %Fe_C %SiO 2_ C Orig. Dup. Orig. Dup. Orig. Dup. Orig. Dup. Nº data Min Max Average Correlation P(Orig. > Dup.) 52.56% 69.23% 53.85% 39.74% Limit P 66.01% 66.01% 66.01% 33.99% T difference Limit T Rel.Diff 0.16% -5.17% 0.09% 4.88%

42 %Fe Conc. DT Dup. %SiO 2 DT Dup. %Fe Head Dup. %Wt rec. Dup. Resource Estimation of the Sheps and Perault Taconite Properties Page 37 Sheps Lake & Perault Lake- DT Duplicates at MRC Sheps Lake & Perault Lake - DT Duplicates at MRC %Fe Head Orig %Wt rec. Orig. Sheps Lake & Perault Lake- DT Duplicates at MRC 75 4 Sheps Lake & Perault Lake- DT Duplicates at MRC %Fe Con. DT Orig %SiO 2 DT Orig. Figure 10.1: Correlation plots of results from NML duplicates

43 Resource Estimation of the Sheps and Perault Taconite Properties Page Data Verification SGS Geostat completed 32 analytical checks of drill core duplicate samples taken from selected NML 2012 diamond drill holes on the Sheps Lake (5 samples) and Perault Lake (5 samples) properties as well as the nearby Howells River North property (not part of this report) as part of the independent data verification program. SGS Geostat also conducted verification of the laboratories analytical certificates and validation of the data set supplied by NML for errors and discrepancies. SGS Geostat considers the data verification done on the Sheps and Perault properties to be current and reliable for resources estimation purposes. During the site visit conducted from August 20 to 22, 2012 by the author, Maxime Dupéré, P.Geo., a total of 32 mineralized core duplicates from the Sheps, Perault and Howells River North (not included in this report) were collected from holes 12PL1030D, 12SL1004D, 12HR1286D, 12HR1035D, 12HR1310D and 12HR1309C by the author and submitted for Davis tube testing and XRF analysis at SGS Minerals laboratory in Lakefield, Ontario, Canada. The duplicate samples were processed using XRF 76C on major oxides. In addition to collecting 32 independent drill core check samples, he checked the location of drill sites with a handheld GPS device SGS independent sampling and assaying During his site visit of August 2012, Maxime Dupéré, P.Geo, collected core samples corresponding to NML original samples. Like for NML duplicate samples, those check samples were made of the remaining half core. Check samples are in 2 holes (5 samples per hole) located in the Sheps Lake and the Perault Lake area. They are between 4.0 and 6.0m long and most of them were taken in the LC unit (5). Check samples were sent to the SGS Lakefield lab for preparation and Davis Tube testing with a protocol similar to that of the Midland Research Center lab. Results from SGS Lakefield related to the four variables reported by MRC (Fe-Head. Wt Rec, Fe_C and SiO 2 _C) are compared to MRC data for the original samples on Figure SGS checked the data using the Sign test and the Student T test (Table 11.1). A bias was observed by SGS Geostat on the Sign test and the Student T test on all of the four variables. In the case of %Fe_Head, it looks like MRC values are almost systematically more than the SGS Lakefield values while for %Wt Rec, MRC values are almost always less than the SGS Lakefield value (points above the diagonal, Figure 11.1). For % Fe_C it looks like MRC values are almost systematically more than the SGS Lakefield values, for %SiO 2_ C it looks like MRC values are almost systematically less than the SGS Lakefield values. These apparent differences between MRC and SGS for the quality of concentrate and Fe Head are somewhat similar to observations on the Lac Ritchie property owned by NML (see NML Technical Report, 2012) and may be linked to the way that quality is measured in the two labs (titration and thermo-gravimetry at MRC and XRF for SGS Lakefield).Although the standard operating procedures from MRC and SGS-Lakefield are similar, some differences might have been unintentionally created from grinding (size reduction) or from calibration of the magnetic separator (Davis Tube) on either laboratory. A detailed investigation of these differences is recommended.

44 Resource Estimation of the Sheps and Perault Taconite Properties Page 39 SGS considers that the number of samples checked (10) is relatively small and conclusions should be interpreted accordingly. SGS recommends increasing the number of check samples in future verifications when more holes are available. Table 11.1: Statistics of SGS check sample data Variable %FeH %Wrec %FeC %SiO 2 C MRC SGS MRC SGS MRC SGS MRC SGS Nº data Min Max Average Correlation P(MRC > SGS) 10.0% 0.0% 100.0% 0.0% Limit P 18.4% 18.4% 81.6% 81.6% T difference Limit T Rel. Diff 3% 7% -1% 55%

45 Resource Estimation of the Sheps and Perault Taconite Properties Page 40 Figure 11.1: Correlation plots of results from SGS check samples

46 Resource Estimation of the Sheps and Perault Taconite Properties Page Mineral Processing and Metallurgical Testing Apart from the Davis Tube testing, there has been no mineral processing or metallurgical test work performed on Sheps & Perault Lakes mineral. However, as Sheps & Perault Lakes mineral is structurally and mineralogically similar to New Millennium Iron s KéMag deposit and the reader can refer to the Technical Report on the Pre-Feasibility Study of the KéMag Iron Ore Project, Québec (BBA, 2009) for related information. With a conventional HPGR + Ball Mill circuit, a processing plant would be expected to produce a concentrate at 2300 Blaine with approximately 1% higher silica than the Davis Tube values. The silica in the magnetic concentrate would thus be approximately 5.5% silica unless a silica cut-off is used to reduce the silica in the feed. The expected weight recovery from the plant would be approximately the same as the Davis Tube weight recovery, if not slightly higher (an increase of 0 to 0.5%). To further reduce the product silica, a flotation circuit could be used. In this case, it is expected that the plant could reach % silica with a loss in weight recovery of 1.5%. Specific test work on Sheps & Perault Lakes the material would be performed at more advanced stages of development.

47 Resource Estimation of the Sheps and Perault Taconite Properties Page Mineral Resource Estimates The resources reported herein are derived from a computerized resource block model. The construction of that model starts with drill hole data, which serve as the basis for the definition of 3D mineralized envelopes with resources limited to the material inside those envelopes. The next step is the selection of drill hole data within the mineralized envelopes in the form of fixed length composites and then the interpolation of the grade of blocks on a regular grid and filling the mineralized envelopes from the grade of composites in the same envelopes. All the interpolated blocks below the overburden/bedrock contact surface make up the mineral inventory and they are classified according to proximity to composites and corresponding precision/confidence level an optimized pit shell was done to verify the validity of the Wt Rec cut-off grade used and however it was not used to restrain the mineral resources and its estimate Drill Hole and Sample Data Sample data used in the construction of the proposed resource model was in drill hole database tables received from NML during SGS site visit and including documents received up to December 18, Data tables used for resource estimation are: A drill hole collar table with collar coordinates, orientations, and lengths of 73 holes totaling m. The effective number of holes is 70 with tags from PL1001D to PL1048D along with sequence SL1001D to SL1022D and totaling m. The other 3 entries are early attempts which had to be aborted (PL1010A 15 m, PL1025D 24 m, SL1006A 4.1 m). The length of effective holes ranges from 40 m to 132 m with an average of 85 m. The majority of holes (48 out of 70) are NQ type with the balance BQ (16) or BTW (19). All holes are vertical. There are no deviation surveys in the drill holes. As illustrated on Figure 13.1 and Figure 13.2, holes are on 24 NE-SW sections separated by 500 m m with spacing of approximately 500 m between holes on the same section. A drill hole assay table with 918 entries. Sample numbers range from 6601 to 6649, 6901 to 7000, 7212 to 7992, 8027 to 8156, and from 9498 to Assay interval length ranges from 0.5 meters to 37 meters with an average length of 5.74 meters. Up to four assay values from Davis Tube tests at Midland Research Center (MRC) are available for each interval : (1) the %Fe of head (Fe_Head) with values from 2.48% to 43.26% (2) the %weight recovery of magnetic concentrate (WtRec) with values from 0 to 47.5% (3) the %Fe of concentrate (Fe_C, missing in 228 samples from lack of concentrate) with values from 60.82% to 72.58% (4) the %SiO 2 of concentrate (SiO 2 _C, missing in 291 samples from lack of concentrate) with values from 0.46% to 6.00%. A total of 24 of the 28 Assay records having lengths over 9 metres had a very poor core recovery and an average %weight recovery of less than 3.5%. A drill hole lithological table with description of 506 lithological intervals (hole name, fromto, litho code) totaling m in the 73 holes. The litho intervals correspond to the stratigraphic units described in section 7 of this report. Table 13.1Table 13.1 shows some statistics on the number and length of the intercepts of holes with those units.

48 Resource Estimation of the Sheps and Perault Taconite Properties Page 43 In addition to the collar, assay and litho drill hole data, regressions of measured S.G. by immersion on similar NML taconite deposits (LabMag and KeMag) have been supplied for various lithological units. Regressions are calculated from numerous samples from each lithological unit (from 20 to 58 samples). Table 13.1: Statistics of litho intervals in drill holes Length (m) Zone Description # Min. Max. Mean BC Black Chert GC Green Chert GRGN Granitic / Gneissic JUIF Jasper Upper Iron Formation LC Lean Chert LIF Lower Iron Formation LRC Lower Red Chert LRGC Lower Red-Green Chert MS Menihek Slate OB Overburden PGC Pink Green Cherty QTE/QTZ Quartzite RS Ruth Slate URC Upper Red Chert There are 55 samples in the LC regression with densities from 3.02 to 3.59 t/m 3 and Iron grades from 19.14% to 37% creating a regression of x %Fe_Head There are 58 samples in the JUIF regression with densities from 3.12 to 3.69 t/m 3 and Iron grades from 23.1% to 38.96% creating a regression of x %Fe_Head There are 20 samples in the GC regression with densities from 2.83 to 3.41 t/m 3 and Iron grades from 8.93% to 34.41% creating a regression of x %Fe_Head There are 49 samples in the URC regression with densities from 3.29 to 3.69 t/m 3 and Iron grades from 26.9% to 40.39% creating a regression of x %Fe_Head There are 51 samples in the PGC regression with densities from 3.19 to 3.64 t/m 3 and Iron grades from 23.19% to 44.04% creating a regression of x %Fe_Head Table 13.2: Summary of SG Regression UNIT # of Samples % Fe SG Min. Max. Ave. Min. Max. Ave. Regression R 2 LC x %Fe_Head JUIF x %Fe_Head GC x %Fe_Head URC x %Fe_Head PGC x %Fe_Head LRC x %Fe_Head LRGC x %Fe_Head

49 Resource Estimation of the Sheps and Perault Taconite Properties Page 44 There are 54 samples in the LRC regression with densities from 3.19 to 3.55 t/m 3 and Iron grades from 24.56% to 36.29% creating a regression of x %Fe_Head There are 53 samples in the LRGC regression with densities from 3.11 to 3.59 t/m 3 and Iron grades from 21.7% to 36.64% creating a regression of x %Fe_Head (Table 13.2) As with many iron ore deposits, the individual density data clearly show a strong correlation of density and %Fe_head grade. It is thus recommended that future density measurements be done on individual samples in addition to composites using water dispersion and pycnometer. Figure 13.1: Map of Sheps drill holes

50 Resource Estimation of the Sheps and Perault Taconite Properties Page 45 Figure 13.2: Map of Perault drill holes 13.2 Outside Limits The only natural boundary for lateral extent of mineralized units at Sheps-Perault is their outcrop on the west side. Since those units have been found in the last drill holes to the east, north and south, limits on those sides are not natural but rather informational limits based on the grid of available drill holes. Given the 500m x 1000m nominal grid of available drill holes, a maximum lateral extent of the resource model can be found by custom contouring points or cells within a 1000 m long radius along N323 and 500 m long short radius along N53 centered on each hole. Furthermore, claim boundaries were considered when creating this custom contour. That contour is called inferred as illustrated on the left side of Figure 13.3 and Figure 13.4 (in blue) and it will limit all potential resources recognized by the drill holes. Since holes are vertical, the contour can be applied vertically i.e. to all benches of the resource block model. Similarly, a another contour was custom fit, defined around points or cells within a 500 m long radius along N323 and short radius 250 m along N53 centered on each hole i.e. the geometrical zone of influence of drill holes on their grid. This contour is dubbed indicated on the right side of

51 Resource Estimation of the Sheps and Perault Taconite Properties Page 46 Figure 13.3 and Figure 13.4 (in red). Section 13.6 further describes why this contour can be used to limit potential indicated resources recognized by the drill hole. The top of the model is the overburden-bedrock surface. It is defined from the supplied topographic surface model. This topography covers most of the Property, with recorded elevations from 477 m to m. Topography data has been re-gridded on the 10m x 10m grid of the resource block model (see below) but only within a contour (in blue on Figure 13.3 and Figure 13.4) drawn about 500 m beyond the contour (in red on Figure 13.3 and Figure 13.4) of maximum resource extension. Within that contour, we have over 970,000 grid points with elevations from 477 to m. Over the same grid points, we interpolate (by simple 2D inverse distance) the overburden thickness recorded in the 73 holes where it varies from 0 m to 9 m with a mean of 2.66 m. A map of interpolated overburden thickness is shown in Figure The elevation of the overburden-bedrock contact surface is defined at each of the grid points as the gridded topography elevation minus the interpolated overburden thickness at that point. The overburden surface was further re-gridded to 50m x 50m to match the geological block model. The bottom of the model is the interpreted footwall of the LRGC layer (see next section). The surface area within the inferred contour but limited to the outcrop line of the LRGC footwall to the west is 42.3 km 2.

52 Resource Estimation of the Sheps and Perault Taconite Properties Page 47 Figure 13.3: Derivation of lateral limits for Sheps resource model

53 Resource Estimation of the Sheps and Perault Taconite Properties Page 48 Figure 13.4: Derivation of lateral limits for Perault resource model

54 Resource Estimation of the Sheps and Perault Taconite Properties Page 49 Figure 13.5: Overburden thickness for Sheps-Perault resource model 13.3 Inside Limits Within top and lateral limits described in the previous section, the proposed Sheps-Perault resource model looks like a multilayered cake with layers corresponding to lithological units identified by NML geologists (see section 7 of this report). Up to 7 layers have been modeled i.e. from top to bottom: LC = lean chert, JUIF = jasper upper iron formation, GC = green chert, URC = upper red chert, PGC = pink-green cherty, LRC = lower red chert and LRGC = lower red green chert. Units below LRGC (i.e. RS, BC and QTE) have not been modeled since they just appear in just a few holes and they do not show any obvious mineralization. Thus the bottom of the resource model is the interpreted footwall surface of the LRGC layer.

55 Resource Estimation of the Sheps and Perault Taconite Properties Page 50 The modeling of contact surfaces between retained layers is based on the interpreted line contacts by NML geologists on drill sections (Figure 13.13). Those lines are discretized at every 50 m and TIN faces connect points on lines of the same nature on two adjacent sections. At both south and north extremities, lines on the last drill section are extruded horizontally along the N323 azimuth. Finally modeled contact surfaces between layers are chopped by the modeled overburden-bedrock contact surface (see previous section). Figure 13.6Figure 13.6 shows the variations of vertical thickness as derived by difference between two successive contact surfaces and for the four most important layers i.e. LC, PGC, URC and LGRC. As a general rule a zero thickness limit corresponds to the interpreted footwall outcrop line. Table 13.3 provides the volumetric characteristics of the proposed resource model i.e. volume, and corresponding average thickness of material in each of the 7 layers of the model. As expected, LC and LRGC are the layers with most material (about 956 Mm 3 and 1422 Mm 3 respectively). Table 13.3: Volumetric of the litho units in the resource model Unit Description Thickness* (m) Volume m 3 Mm 3 LC Lean Chert ,078, JUIF Jasper Upper Iron Formation ,345, GC Green Chert ,740, URC Upper Red Chert ,222, PGC Pink Green Cherty ,370, LRC Lower Red Chert ,596, LRGC Lower Red-Green Chert ,422,039, * Calculated from the average thickness of mineralized intervals TOTALS 2,831,393, ,831.39

56 Resource Estimation of the Sheps and Perault Taconite Properties Page 51 Figure 13.6: Thickness of mineralized units from modeled contact surfaces 13.4 Compositing, and Statistical Analysis Statistics of sample length and available values by litho unit are in Table Most of samples are in the three units LC, PGC, and LRGC however all seven units are utilized in the block grade interpolation. The %Fe of head averages about 30% in JUIF, URC, LRGC, and PGC (28.5%) but it is lower in the LC marker horizon (25%) and much lower in the GC marker horizon (18% average) and higher in the LRC unit (average 34%). Average recovery is high in LRC (29%) and JUIF (26%), moderate in LRGC (22.4%), URC (22.3%), PGC (21.4%) and LC (18.3%) and expectedly low in GC (3.9%). The quality of the magnetic concentrate is about the same in all units, even in those with very low recoveries. For samples in all zone presented, the average %Fe keeps around 69%-71% and the average %SiO 2 between 1.3% and 2%.

57 Resource Estimation of the Sheps and Perault Taconite Properties Page 52 Although the majority of original samples are 6 m long, there exists both some shorter and some longer samples. Therefore, statistics might be biased by some high or low values measured on short or long intervals. The standardization of sample size is done by numerical compositing. The most natural composite length is obviously the 6 m of a majority of samples. In order not to drop any assay information, each drill hole intercept with a litho unit is split into composites of equal length the closest to 6 m. The %Fe of head and %Weight recovery of composites are standard averages weighted by length while the %Fe and %SiO 2 of concentrate are averages weighted by both length and recovery of the sample fractions within the limits of composites. In the thin GC unit, the final composites are actually the original samples. Statistics of composite data in the seven units of interest are on Table They are consistent with statistics of sample data in the same unit. Histograms of composite grades in the same seven units are on Figure 13.7 to Figure As noticed before, the variability of %Fe head is quite limited around 30% Fe for PGC and LRGC, 25% Fe for LC and 35%Fe for URC, LRC and JUIF with just a few low outliers in URC and JUIF. GC has a distribution around 20% and is wide for the sample population, with some highs and lows. Coefficients of variation (i.e. standard deviation divided by mean) do not exceed 22% except in the thin GC unit. Recovery has a high variability; most units have a coefficient of variability above 40%. Despite significant differences of mean %Fe head and weight recovery, %Fe and %SiO 2 of concentrate are rather uniform in all units with typical negatively skewed (mostly high with a tail of low value) histograms for %Fe and corresponding positively skewed (mostly low with a tail of high values) histograms for %SiO 2. The presence of two relatively distinct populations in the Wt Rec histograms (mostly in the LC and LRGC units) can be explained by the relative difference between the Sheps property (medium to high recovery) with the Perault property (low recovery). Correlations between composite values are about the same in all units. They are illustrated on Figure 20 for unit LC. There is no useable correlation (R=+0.4) between recovery and %Fe of head nor between %Fe of head and %Fe of concentrate (R=+0.2). There no correlation between %Fe of concentrate and the %weight recovery thus indicating the quality of iron concentrate has no bearing on the weight recovery. The really strong (and expected) correlation is the negative one (R= to ) between the %Fe and the %SiO 2 of concentrate, except for the URC unit (R=0.287) however there is a limited range of high and low values.

58 Resource Estimation of the Sheps and Perault Taconite Properties Page 53 Table 13.4: Statistics of sample data according to litho units ZONE LENGTH %Fe Head %Wrec %Fe Conc. %SiO 2 Conc. # Min Max Mean # Min Max Mean # Min Max Mean # Min Max Mean # Min Max Mean LC JUIF GC URC PGC LRC LRGC TOTAL Table 13.5: Statistics of composites according to lithological units LENGTH %Fe Head %Wrec %Fe Conc. %SiO 2 Conc. ZONE # Min Max Mean # Min Max Mean # Min Max Mean # Min Max Mean # Min Max Mean LC JUIF GC URC PGC LRC LRGC TOTAL

59 Resource Estimation of the Sheps and Perault Taconite Properties Page 54 Figure 13.7: Histogram of % Fe Head for composites of various lithological units

60 Resource Estimation of the Sheps and Perault Taconite Properties Page 55 Figure 13.8: Histogram of %Wt Rec Head for composites of various lithological units

61 Resource Estimation of the Sheps and Perault Taconite Properties Page 56 Figure 13.9: Histogram of %Fe concentrate for composites of various lithological units

62 Resource Estimation of the Sheps and Perault Taconite Properties Page 57 Figure 13.10: Histogram of %SiO 2 Concentrate for composites of various lithological units

63 Resource Estimation of the Sheps and Perault Taconite Properties Page 58 Figure 13.11: Correlation of 6 m Composite values in the LC unit 13.5 Block Grade Interpolation The four quality parameters analyzed in the previous section are interpolated in blocks on a regular grid. That grid is oriented parallel to the drill sections i.e. columns are numbered along a rotated x axis along the N53 azimuth and rows along a rotated y axis along the N323 azimuth. The origin of the grid i.e. the center of the block in the first column (numbered from west to east) and first row (numbered from south to north) has the following UTM coordinates: 635,300 me and 6,050,650 mn. The selected block size is 25 m along the rotated x and 50 m along the rotated y. That size is obviously tiny when compared to the drill hole grid of 500 m x 1000 m. The main reason to choose blocks with such a small size is to better reflect the geometry of litho units slowly dipping to the NE. It is also the size of blocks used in the resource modeling of the LabMag and KéMag deposits further to the south which both belong to NML see Geostat (2005) and Geostat (2007a). Vertically, blocks are 15 m high. Although the bench heights previously used for the KéMag and LabMag Pre-Feasibility studies was 13 m, the most recent work by mining consultant Met-Chem on those projects has suggested that a bench height of 15 m would be more suitable. All together, the mineralized material limited as described in previous sections 14-2 and 14-3 is covered by blocks in up to 97 columns, 500 rows and 23 benches from elevations z=270 to z=600.

64 Resource Estimation of the Sheps and Perault Taconite Properties Page 59 In a first pass, the volumetric fraction is calculated in blocks of each of the seven litho units above the footwall of LRGC i.e. LC, JUIF, GC, URC, PGC, LRC, and LRGC. Then the four quality parameters of block fractions in these units are interpolated from composite data in the same unit. Interpolation is done by inverse distance squared (ISD2). With sample data on a regular grid (no sample clustering in high grade), low nugget effects and long ranges, ISD2 is known to provide block estimates very similar to ordinary kriging (OK). The basic search ellipsoid is a flat 1200m x 600m x 50m tilted by 10 to the N53 (a review of interpreted litho units on sections shows this 10 dip angle for most on them). The 1200 m x 600 m elliptic outline on sub-horizontal planes is designed to capture composites from at least 4 neighbor holes on the 1000x500 m nominal grid. In a first interpolation run, we need a minimum of 5 composites (3 in LRC, URC and the thin GC unit) in a minimum of 3 drill holes (maximum number of composites from the same drill hole is 2) within the 1200 m x 600 m x 50 m ellipsoid for allowing the block interpolation in that run. The maximum number of composites retained is 25 (15 in LRC, URC and the thin GC unit). In LC, 28% of the blocks are interpolated in that first run. Blocks that do not meet the minimum number of composites and holes on the first run are tested in a second run with a 2400 m x 1200 m x 100 m ellipsoid of similar orientation. Again, interpolation takes place if there are at least 5 composites (3 in LRC, URC and GC) in at least 3 drill holes within the larger ellipsoid. The maximum number of composites is raised to 30 (20 in LRC, URC and GC). In LC, 56.7% of the blocks are interpolated in that second run. The few blocks that cannot be interpolated in the second run are interpolated in a third and last run with a 4800 m x 2400 m x 200 m ellipsoid and a minimum number of composites and holes being one. The third pass for the thin GC unit used a 9600m x4800m x 400m ellipsoid of similar orientation, furthermore a fourth pass was applied to the GC unit for SiO 2 _C only using a 24000m x 12000m x 1000m ellipsoid. A complete listing of the percent blocks interpreted is summarized in Table 13.6, and a listing of search parameters is found in Table 13.7 and Table Table 13.6: Summary of blocks classified by pass Pass 1 Pass 2 Pass 3 Pass 4 Total ZONE Blocks Percent Blocks Percent Blocks Percent Blocks Percent Blocks Percent LC % % % % JUIF % % % % GC % % % % % URC % % % % PGC % % % % LRC % % % % LRGC % % % %

65 Resource Estimation of the Sheps and Perault Taconite Properties Page 60 Table 13.7: Estimation settings search parameters Estimation Method Unit LC JUIF GC URC PGC LRC LRGC Search Ellipse Estimation Run Inverse Distance Squared (ID2) Maximu minimum m Nb of Number of Samples Samples Max Nb of Samples per Hole Pass 1 # Pass 2 # Pass 3 # Pass 1 # Pass 2 # Pass 3 # Pass 1 # Pass 2 # Pass 3 GC # Pass 4 GC # Pass 1 # Pass 2 # Pass 3 # Pass 1 # Pass 2 # Pass 3 # Pass 1 # Pass 2 # Pass 3 # Pass 1 # Pass 2 # Pass 3 GC # Table 13.8: Estimation settings ellipse parameters Search Ellipse Avimuth Dip Spin Major Axis (m) Medium Axis (m) Minor Axis (m) Pass Pass Pass Pass 3 GC Pass 4 GC Two of the interpolated parameters, %Fe_head and %Wt Rec are additive variables and can be interpolated as weighted average of sample data. The other two variables, %Fe and %SiO 2 of concentrate are non additive since averages need to be weighted by weight recovery. Theoretically, one should interpolate the product of concentrate grade by recovery and divide the interpolated product by the interpolated recovery in order to get interpolated grade of concentrate. The problem with this theoretically sound approach is with missing concentrate grades, particularly %SiO 2 : since we do not interpolate the products and recovery with the same composites, the final ratio of interpolated values may yield odd results. Hence the theoretically incorrect direct interpolation of %Fe and %SiO 2 of concentrate is preferable since it provides more robust results.

66 Resource Estimation of the Sheps and Perault Taconite Properties Page 61 Statistics of interpolated block values are on Table As expected from the regular grid of composites, average block values are close to average composite values in the same unit (Table 13.5). As expected too, the range of block values is slightly narrower than the range of composite data. Block estimates are illustrated by those of the LRGC unit in the twelfth bench on Figure Quality estimates for unit fractions in the same blocks can be merged together for a total fraction of LC+JUIF+GC+URC+PGC+LRC+LGRC in a single set of quality estimates of that fraction in the block. In that operation, quality estimates of each unit fraction are weighted by the volumetric unit fraction and also the calculated density based on iron content. For detailed density information refer to section 13.1 and Table All together, we have 185,152 blocks 50x25x15 m in the seven mineralized units LC+JUIF+GC+URC+PGC+LRC+LRGC. Statistics of estimated quality values for those total fractions in blocks appear on the last line of Table Drill sections through the block model are on Figure

67 Resource Estimation of the Sheps and Perault Taconite Properties Page 62 Table 13.9: Statistics of interpolated block values according to litho units Zone # of blocks Volume Fraction of block %FeH %Wrec %FeC %SiO 2 C (Mm 3 ) Min Max Ave Min Max Ave Min Max Ave Min Max Ave Min Max Ave LC 68, JUIF 17, GC 15, URC 16, PGC 25, LRC 6, LRGC 102,587 1, All 185,152 2, The all is a summary of the complete block model, since blocks are shared between units, the summations, minimum, maximum and average values are taken from a complete list of interpolated blocks.

68 Resource Estimation of the Sheps and Perault Taconite Properties Page 63 Figure 13.12: Block estimates in a bench for LRGC unit

69 Resource Estimation of the Sheps and Perault Taconite Properties Page 64 Figure 13.13: Drill Sections through Block Model

70 Resource Estimation of the Sheps and Perault Taconite Properties Page Block Categorization and Mineral Inventory The Sheps-Perault magnetite bearing taconite formations are currently recognized by 73 vertical drill holes on a grid of 1km by 0.5 km, 3 of which were abandoned. Those holes cover a NW-SE area with a strong magnetic anomaly recognized by the Fugro airborne survey of This area is bounded to the NW and SE by two major fault zones easily recognizable from the survey. Aside from those two bounding fault zones, the map of the magnetic anomaly (Figure 8.1) does not show any major discontinuity. The geological continuity of the mineralized units has been demonstrated by the results from the 70 holes. In most holes with the occasional disappearance of the marker horizon of the thinner units (predominantly GC, URC, PGC and LRC), the stratigraphic sequence of (from top to bottom) LC + JUIF + GC + URC + PGC + LRC + LRGC can be recognized with similar thickness data for all intercepts in the same unit. That stratigraphic sequencing is not arbitrary since it is supported by a mineral signature particular to each unit i.e. medium Fe + med magnetite in LC, high Fe + high magnetite in JUIF, low Fe + low magnetite in GC, high Fe + medium-high magnetite in URC, medium Fe + medium-high magnetite in PGC, high Fe + high magnetite in LRC, high Fe + medium-high magnetite in LGRC. Additionally, there is a difference between the southwestern part of the Perault property area and the more northwestern part corresponding to the Sheps property area. Despite the long spacing between holes, the grade continuity is demonstrated by the low variability of the % Fe Head data as well as the geological information and interpretations from NML. There is however, an observed variability in the recovery data that can be explained by the difference between the Sheps (higher recovery) and the Perault (lower recovery) property areas. Additional drilling and/or geological data gathering and interpretation would be necessary in order to ascertain this difference. Given the well documented geological interpretation of the area by the Company and; the relative low variability of the Fe_Head data, as well as the overall continuity of the mineralization between sections and the 70 holes on the 1000 m x 500 m grid, we can classify the mineralized material recognized by those holes in the indicated category. Hence the mineral inventory made of blocks within the 1000 m x500 m cells of influence of holes (the red contour of Figure 13.3 and Figure 13.4) is classified as indicated. The balance of the mineral inventory (between the red contour and the blue contour on Figure 13.3 and Figure 13.4) is classified in the inferred category. A second drilling campaign is warranted in order to confirm with higher certainty the geological and grade continuity that allowed classifying the material in the indicated category and upgrade some of the resources to the measured category. The mineral inventory is given at various cut-offs of weight recovery (Table 13.10). Those cut-offs are applied to the estimated weight recovery of the block fraction made of LC + JUIF + GC + URC + PGC + LRC + LRGC units.

71 Resource Estimation of the Sheps and Perault Taconite Properties Page 66 The variations of ROM tonnage and head (%Fe, %Weight recovery) plus concentrate (%Fe, %SiO 2 ) quality with cut-off are graphed on Figure The decrease of tonnage with cut-off is moderate at low cut-off (0-25% weight recovery) as most of the material in all of the units is eliminated. Then the rate of decrease is accelerated for high cut-offs (25-28% weight recovery) as the remainder of the material is eliminated, mostly in the marginal material in the URC, LRC and the other units is progressively eliminated. It decreases again for high cut-offs (28-30% weight recovery as the bulk of the JUIF and LRC indicated mineralization is touched upon. The increase of average weight recovery with cut-off follows the same trends with a small advantage for indicated material for low and medium cut-offs.

72 Resource Estimation of the Sheps and Perault Taconite Properties Page 67 Table 13.10: Mineral Inventory of Sheps and Perault at various weight recovery Cut-Offs Estimated Resources Sheps &Perault COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % 0 Indicated 1,715 5, Inferred 1,117 3, Indicated 1,656 5, Inferred 1,035 3, Indicated 1,608 5, Inferred 935 2, Indicated 1,559 5, Inferred 805 2, Indicated 1,505 5, Inferred 681 2, Indicated 1,434 4, Inferred 578 1, Indicated 1,342 4, Inferred 438 1, Indicated 1,253 4, Inferred 322 1, Indicated 1,208 4, Inferred 301 1, Indicated 1,163 3, Inferred Indicated 1,114 3, Inferred Indicated 1,057 3, Inferred Indicated 942 3, Inferred Indicated 811 2, Inferred Indicated 609 2, Inferred Indicated Inferred Indicated Inferred Indicated Inferred

73 %Fe concentrate above cut-off %SiO2 concentrate above cut-off Tonnage ROM above cut-off %FeH and %Weight recovery above cut-off Resource Estimation of the Sheps and Perault Taconite Properties Page 68 Sheps-Perault Mineral Inventory %Weight Recovery Cut-off Indicated Tonnage Inferred tonnage Indicated WRec Inferred WRec Indicated FeH Inferred FeH Sheps-Perault Mineral Inventory %Weight Recovery Cut-off Indicated FeC Inferred FeC Indicated SiO2C Inferred SiO2C Figure 13.14: Mineral Inventory with Weight Recovery Cut-Off As expected from the mild positive correlation of %Fe Head and Weight recovery in samples (Figure 13.11), there is a slight increase of average %Fe of head with the cut-off, with a slight advantage for the inferred material. The mild negative correlation of %Fe of concentrate and weight

74 Resource Estimation of the Sheps and Perault Taconite Properties Page 69 recovery in samples (Figure 13.11) explains the slight but steady decrease of %Fe of concentrate and increase of %SiO2 of concentrate with the cut-off Estimated Resources Resources are that part of the mineral inventory with a reasonable prospect of economic extraction which means both a cut-off grade and an optimized pit shell i.e. blocks with estimated values above a cut-off and within the limits of an optimized pit shell. Traditionally, the cut-off used to report resources in the taconite deposits of the Labrador Trough is a minimum 18% weight recovery of the magnetic concentrate from Davis Tube test on material ground to 325 mesh (see WGM (2006), Geostat (2007a+b) and BBA (2009)). According to BBA (2009) the unit cost and concentrate values used in the PFS of KéMag suggest a lower marginal cut-off (such that the concentrate value pays for the processing cost to produce that concentrate). In that study, a pit optimization is run with a C$4.03/t ROM total processing + G&A cost and a concentrate value of C$49.92/t CC (unit mining cost is C$1.75/t ROM) hence a marginal weight recovery cut-off of : 4.03/49.92 = 8%. For Sheps and Perault, NML is currently proposing the following tentative parameters: concentrate value = $68.41/t CC, crushing and concentration = $11.45/t CC, concentrate handling (pipeline + filtration + port/loading) + G&A = $3.21/t CC, mining cost (ore and waste) = $2.50/t ROM and mining cost (overburden) = $1.70/t OVBD. Given that the crushing and concentration cost is given for an average weight recovery of 28% (above the traditional COG of 18%), it translates into a crushing + concentration cost of: 11.45*0.28 = $3.21/t ROM hence a marginal cut-off of: 3.21/ = 4.69%. We have run a Whittle optimized pit shell based on the mineral inventory described in previous sections (both indicated and inferred blocks), the above NML cost and value parameters and maximum pit slope angles of 50. The optimized pit shell (Figure 13.15) includes most of the mineral inventory and all the blocks with an estimated weight recovery above 15% recovery cut-off. As a result, the proposed resources for Sheps-Perault are made of all the blocks of the mineral inventory above 15% minimum recovery cut-off (Not restrained by any pit shell). They are listed by unit, total and total after merging unit fractions in blocks in Table As expected from the statistics of sample data in each unit, there are no resources above 15% recovery in the GC. The 0 tonnage in resources of the indicated GC unit in the following tables corresponds to a tonnage less than 500,000 tonnes and is therefore described as 0.

75 Resource Estimation of the Sheps and Perault Taconite Properties Page 70 Table 13.11: Estimated Resources of Sheps & Perault by Unit and Total Estimated Resources Sheps & Perault Unit COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C Number WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % of Blocks LC 15 Indicated ,254 LC 15 Inferred ,570 JUIF 15 Indicated ,924 JUIF 15 Inferred ,087 GC 15 Indicated GC 15 Inferred URC 15 Indicated ,548 URC 15 Inferred ,474 PGC 15 Indicated ,038 PGC 15 Inferred ,107 LRC 15 Indicated ,602 LRC 15 Inferred ,175 LRGC 15 Indicated 765 2, ,011 LRGC 15 Inferred ,836 TOTAL 15 Indicated 1,206 4, ,384 TOTAL 15 Inferred 314 1, ,249 FINAL 15 Indicated 1,208 4, ,199 FINAL 15 Inferred 301 1, ,228 Dated February 7th, Mineral resources are not mineral reserves and do not have demonstrated economic viability. Bulk density was respectively attributed according to linear regression formulas on Fe Head for each unit total= Sum of unit resources. Final=Cut-off applied to weight recovery of the LC, JUIF, GC+URC+PGC+LRc+LRGC in each block The 0 tonnage in resources of the indicated GC unit in the table above corresponds to a tonnage less than 500,000 tonnes and is therefore described as 0.

76 Resource Estimation of the Sheps and Perault Taconite Properties Page 71 Table 13.12: Estimated Resources of Sheps by Unit and Total Estimated Resources Sheps Unit COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C Number WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % of Blocks LC 15 Indicated 1,477 4, LC 15 Inferred JUIF 15 Indicated 292 1, JUIF 15 Inferred GC 15 Indicated GC 15 Inferred LRC 15 Indicated LRC 15 Inferred LRGC 15 Indicated 342 1, LRGC 15 Inferred PGC 15 Indicated PGC 15 Inferred URC 15 Indicated 3,277 11, URC 15 Inferred 512 1, TOTAL 15 Indicated 587 2, TOTAL 15 Inferred FINAL 15 Indicated 596 2, FINAL 15 Inferred Dated February 7th, Mineral resources are not mineral reserves and do not have demonstrated economic viability. Bulk density was respectively attributed according to linear regression formulas on Fe Head for each unit total= Sum of unit resources. Final=Cut-off applied to weight recovery of the LC, JUIF, GC+URC+PGC+LRc+LRGC in each block The 0 tonnage in resources of the indicated GC unit in the tables above and below corresponds to a tonnage less than 500,000 tonnes and is therefore described as 0. Table 13.13: Estimated Resources of Perault by Unit and Total Estimated Resources Perault Unit COG Category Volume Tonnage Density Fe_Head WtRec Fe_C SiO2_C Number WtRec(%) (Mm 3 ) (Mt) (t/m 3 ) % % % % of Blocks LC 15 Indicated 988 3, LC 15 Inferred 490 1, JUIF 15 Indicated JUIF 15 Inferred GC 15 Indicated GC 15 Inferred LRC 15 Indicated LRC 15 Inferred LRGC 15 Indicated LRGC 15 Inferred PGC 15 Indicated PGC 15 Inferred URC 15 Indicated 4,372 14, URC 15 Inferred 1,400 4, TOTAL 15 Indicated 619 2, TOTAL 15 Inferred FINAL 15 Indicated 613 2, FINAL 15 Inferred Dated February 7th, Mineral resources are not mineral reserves and do not have demonstrated economic viability. Bulk density was respectively attributed according to linear regression formulas on Fe Head for each unit total= Sum of unit resources. Final=Cut-off applied to weight recovery of the LC, JUIF, GC+URC+PGC+LRc+LRGC in each block

77 Resource Estimation of the Sheps and Perault Taconite Properties Page 72 Perspective view with no iron cut-off inside the shell Perspective view with iron-cut-off greater than 18% Perspective view with iron-cut-off greater than 15% not restricted to the shell BASE CASE Figure 13.15: Perspective view of optimized pit shell

78 Resource Estimation of the Sheps and Perault Taconite Properties Page 73 On Figure the View is to the northeast. Top: blocks within the pit shell with no cut-off grade, Middle: Blocks within pit shell and above cut-off grade of 18%. Bottom: Blocks within pit shell and above cut-off grade of 15%. Blocks are coloured according to iron head grade. Limits of inferred in blue, and claims in black are shown.

79 Resource Estimation of the Sheps and Perault Taconite Properties Page Adjacent Properties To the north of the Perault Lake and Sheps Lake Properties and along the Millennium Iron Range, NML holds the Lac Helluva Property which is contiguous to the KéMag and LabMag Properties further to the northwest. On the KéMag Property, based on 88 holes drilled in and totaling 10,774 m, NML has identified a resource of 2,448 Mt measured +indicated at 31.3% Fe of head, 26.3%Weight recovery, 69.4% Fe and 2.7% SiO 2 of concentrate and 1,014 Mt inferred of similar quality, both above the traditional cut-off of 18%Weight recovery see Geostat (2008). On the LabMag Property, the current resource estimate from Geostat (2007b) is 43,665Mt measured + indicated at 29.6 Fe of head, 26.4%Weight recovery and 70.0%Fe and 2.2%SiO 2 of concentrate plus 1,475 Mt inferred of similar quality, both above the traditional cut-off of 18%Weight recovery. That estimate was derived from 216 holes drilled from 2004 to 2006 and totaling about 17,576.4 m. Both resources have been the subject of PFS studies (see WGM, 2006 and BBA, 2009). Metallurgical testing and pilot scale studies for producing concentrate and pellets were conducted in Germany. LabMag and KéMag deposits contain a combined total of 5.6 billion tonnes of proven and probable reserves at 30.24% Fe, 1.3 billion tonnes of measured and indicated resources at 29.92% Fe and 2.2 billion tonnes of inferred resources at 30.16% Fe. NML and Tata Steel Minerals Canada Ltd. ( TSMC ) engaged SNC-Lavalin, the Engineering Consulting Co. to carry out a feasibility study for the taconite project. Lac Otelnuk Property of Adriana Resources Inc is situated approximately 155 km in a straight line northwest of the village of Schefferville near the border with Labrador, and 225 km south of the village of Kuujjuaq. The resource estimate by Watts, Griffis and Mc Ouat Limited (WGM) in August 2012 amounting to 5.84 billion tonnes of Indicated Resource grading 28.7.% T Fe and billion tonnes of Inferred, Resources at 30.4% T Fe and 29.5 billion tonnes of Measured Resource grading 5.51% (available on SEDAR). The Hayot Lake project, part of the Attikamagen iron property is located in northeastern Québec, approximately 22 kilometres north of the town of Schefferville; 220 kilometres north of Labrador City, Newfoundland and Labrador; and 500 kilometres north of Sept-Îles, Québec. Labec Century Iron Ore Inc. (Labec Century), a subsidiary of Century Iron Mines Corp. (Century), executed an agreement with Champion Minerals Inc. (Champion); Labec Century holds 56% interest on the property and Champion Minerals Inc. (Champion) 44%. In November 2012 Labec Century published a mineral Resource Evaluation about Hayot Lake taconite Iron project, the inferred resource is 31.25% Fe (report available on the website). There are companies like Labrador Iron Mines Limited which exploits DSO around properties. A map of the adjacent properties surrounding and/or near the Sheps and Perault properties is available in Figure 14.1.

80 Resource Estimation of the Sheps and Perault Taconite Properties Page 75 Figure 14.1: Map of adjacent Properties