SPECTRUM ENGINEERING To: Jennifer MDNR From: Donald Sutton Date: February 24, 2012 Subject: HydroMet and Stockpiles - review of Barr responses to comments PolyMet Geoteehnieal Modeling Work Plan comments Flotation Tailings Basin Geoteehnical Model for SDEIS, FEIS and Permitting: I, Item 3a. The bentonite seal is a hail Mary type of concept in my opinion. I believe it will exacerbate erosion and slope failure and will eventually fail, so I recommend that the stability analysis should assume the bentonite doesn?t prevent seepage so far as stability is concerned. a. What is the stability of the side slopes if a layer of tailings is placed above the bentonite and it becomes saturated? Will the bentonite slope fail? b. What if the bentonite slope is saturated and there is an earthquake or a thunderstorm? 2. I am concerned about long term climate change and how it will affect the water balance in the tailings facility because this can affect the water level, the water head, the saturation, and most importantly, erosion. Erosion can cause the shape of the embankments to change over time, it can cause erosion and gullying that can create a pathway for water to escape from the pond. This bothers me, because the proposed design is temporary and will fail unless it is perpetually maintained. I am surprised that the Minnesota statutes allow a temporary impoundment structure to be permitted permanently. This wouldn?t be allowed in otherjurisdictions. I realize that this will be addressed during the permitting and ?nancial assurance review. Estimating the liabilities will be contentious. Stockpile Geoteehnical Models for SDEIS, FEIS and Permitting: 1. Is there a schedule for collecting the foundation data? How does it relate to the EIS if the data isn't obtained in time? a. As a practical matter, I think the stockpile geotechnical designs pose no geotechnical EIS related concerns provided they honor the statutes. My concerns are related to water management, erosion control and geochemical issues. b. There is a potential minor leak risk with the liners, but this is minimal and has been addressed. I would like to see the sub -base designed so that any leakage can be collected or directed to the pit pumping system. At some point, someone needs to check this, but it isn?t a geotechnical item. 14134'h Avenue North - Billings, Montana 59101 406534-4660 - Fax406f259-1456 MCEA Comments EX. 23 UPDATE DRAFT VERSION 1/31/12 GeotechnicaliGeochemical Questions Related to Polymet Tailingsl Dana. DosteIt - January 18, 20 2 PolyMet Mining proposes building a new tailings storage facility on top of the existing LTV tailings basin. Poly-Met plans on mining old LTV coarse tailings for construction materials to be used in the perimeter dams that will contain the PolyMet tailings between the perimeter dams and above the lower lying LTV tailings. PolyMet ?irther proposes to place a bentonite cap on top of the tailings and maintain a wet cover to reduce or prevent oxygen from entering the PolyMet tailings. Poly-Met plans on using the upstream construction method, which will ultimately result in the perimeter dams being constructed over the PolyMet tailings. Dam Safety has experience with tailings dams that are constructed from the residue from the taeonite industry, but has no experience dealing with railings that will be derived from minerals in the Duluth Complex. Furthermore, Dam Safety has numerous concerns with this project because the tailings dams must Function properly for an extended period of time - we?ve heard on the order of 900 years. Our ?rst concern is whether the PolyMet tailings will form a structurally sound base to support the perimeter dams. Our second concern is that the proposed wet cap will signi?cantly increase the potential for a dam failure, and will result in costly monitoring and maintenance over the life of the project. (Including monitoring costs to DNR for 900 years) Questions: 1. Geochemical: Will the PolyMet tailings be geochemically stable over the life of the tailings basin (900 years), or will they degrade into secondary minerals? Will these minerals migrate? Can we expect deformations within the basin due to the chemical breakdown and possible migration of the PolyMet tailings? What happens if the bentonite cap leaks and water enters the tailings? I LAM Technical Input: The tailings are proj ceted to have the following primary mineral percentage ranges: 50-80% plagioclase, 10-15% olivine, 4-594: elinopyroxene and <1 0.5% sul?des. Natural weathering processes will lead to breakdown of the primary minerals into less competent secondary clay and carbonate minerals. However, the extent and rate to which these reaction processes would occur under this setting is not clear. Leaking through the bentonite cap would likely increase the extent and rate of these weathering processes. 2. Physical: Will the tailings be physically stable over the life of the tailings basin? Will they be prone to rolling, crushing andfor collapse? Deteriorate due to Freeze! thaw? Fracture due to moisture? it is our understanding that the PolyMet tailings will be rounder, less jagged and less likely to interlock Is this correct? 1 Supplemented by LAM technical input provided by Zach Wenz; EIS Guidance proposed by Stuart Arkie?f UPDATE DRAFT VERSION 1/31/12 COE Technical Input: Slope stabilitv of the TB darn is the principle mode of failure associated with these questions. From that standpoint there are two main properties that are considered with particle crushing or particle deterioration: shear strength and penneabilitv. The follow on question is whether these phvsical changes from crushing or deterioration will affect the material structure leading to a greater potential for collapse from a looser state to a denser state. Lade. et.al.- discusses factors affecting particle breakage in an ASCE paper entitled ?Signi?cance of Particle Crushing in Granular Materials." Some of these include stress level. where the higher the stress (especiallv shear stress) the greater the breakage; particle size as the larger the size the greater the number of ?aws or defects (probablv not an issue on Polvmet]; particle angularitv since the more angular the more fracturing at particle points or breakage along narrow dimensions; unifonnitv of materials as the more uniform the less the number of contracts and therefore higher contact stresses: and mineral hardness where the harder and stronger particles have lesser breakage for the same stress. Assuming the hardness of Plggioelase is about that of Orthoelase. the hardness is about 70% of quartz. Taleb Al-Rousan- et.al. reports in the paper ?Effect of Inherent Anisotropv on Shear Strength Following Crushing of Natural Aqaba Subgrade Sand? that coarse granitic sand particles with 2.8 diameter experience breakage at a pressure of (4 [-770 psi). Using a moist densitv of 100 and no groundwater conditions this would equate to a ?ll height of 413 ft. This is a compressive stress and savs nothing about breakage under shearing but based on information reviewed it is felt that particle crushing and degradation isn't a signi?cant concem. Instead. the structure or fabric of the tailings and their potential to ligucfv under dvnamic or static loading is the governing behavior that should be investigated. Freeze-thaw degradation might have an effect on the outer band of the TB (sav the outer 7-l 0 ft) but since processing has eliminated bedding plane and fracture defects we are reallv only talking about degrading minerals and from judgment it is expected that this occur very slowlv. According to Lade and others. less angular particles are subicet to less breakage and therefore less potential change in shear strength and decrease in penneabilitv from design conditions. 3. Alterations: Will Polymet tailings congeal overtime (900 years) due to pressure and the presence of Ca and Mg, or will they remain loose with characteristics similar to a fine sugar or flour. CUE Technical lnput: Aging of soils will likelv have some impact on shear strength. As water levels drop and effective stresses increase it 1navbe expected that the grains will move into a denser. more stable Dosition. This doesn't mean that the soils are no longer prone to liquefaction. It is felt that once steadv state conditions are reached and no further consolidation occurs the aging process can continue and include cementation effects. Cementation can take place over hundreds or thousands of vears and it seems to be uneonservative to count on this cementation or congealing to allow the use of higher shear for design. 4. Is there a geochemical concern if there is a darn failure and PolyMet tailings spill into the surrounding wetlands, so long as no one gets hurt, or are there no eoneems?.? US EPA has stated they would require the tailings to be picked up. LAM Technical Input: PolyMet has projected the sulfur content of the tailings to be in the neighborhood ofll. 1% to Given the small particle size ofthe tailings and their likely widespread deposition resulting from a dam failure it is likely that water quality violations regarding sulfate, metals, or suspended solids could arise. UPDATE DRAFT VERSION 1/31/12 5. Are there any other geochemical, or other concerns related to the tailings that could impact dam stability over the life (900 years) of the project that we should be aware of? Are the concerns we have expressed valid? 1- LAM Technical Input: The potential for PolyMet tailings to breakdown into less competent minerals (clays and carbonates) overtime could negatively impact the stability of the tailings basin, However, the probability that this would lead to failure of the tailings basin is not clear. Therefore, requesting a projection of the effect of mineral weathering to the stability of the tailings basin is valid. Document Item #/page General PTM 10.1.3.2 Waste Rock and Ore, p188 10.1.3.2 Waste Rock and Ore, p189 10.1.3.2 Waste Rock and Ore, p189 10.1.3.3 Flotation Tailings, p 190 10.1.3.3 Flotation Tailings, p 190 Comment Section 10 Mine Waste Characterization Appendix 2, 4, 6, 7 and 17 The accuracy of the geochemical and pilot plant testing conclusions depends on how accurately the samples represent the mass of ore and waste. I don’t know how the samples were collected, etc. so can’t comment. However, keep in mind that there is always sampling error, and the sample results don’t always scale up perfectly. This is a good starting point for fine tuning later. Location of Tables. The text would be easier to follow if the tables were inserted where referenced, rather than at the end. Of the 18,800 waste rock and ore samples used to characterize the ore and waste, apparently only 82 samples were used to characterize the waste. This makes me wonder how much confidence we have when scaling up to the full 358 million tons of waste that will be mined, and the 218 million tons that will be stockpiled. As a practical matter, if PolyMet conducts a rigorous blast hole sampling program and uses the results to ensure the waste is properly managed, this may not matter. However, if the small sample size does not accurately reflect the projected water chemistry, it may influence the water treatment costs. This is just one more reason to maintain a healthy contingency on the water treatment costs. I haven’t found the proposed day-to-day blast hole sampling and waste ore digging plan that ensures that the waste rock will be segregated by %S. Need a QA/QC plan for doing this, and then an independent way to sample and test Cat1 to verify. Then, need a plan to address non-compliance. Cat 1 waste rock. Some may be used for construction. Will not generate acid, but “may” release metals. Can the Cat 1 material be segregated into two categories: Cat1a will NOT release metals and Cat1b MAY release metals? I get the impression that insufficient waste sampling was completed to answer this question, but have not yet scoured the 3D sample data set to form an opinion about the representativeness. Did PM provide a 3Ddataset? Hypothetically, the size and cost of the Cat1b (MAY release metals) pile could be reduced, and the costs to reclaim the pile could be released id the Cat1a (will NOT release metals) Flotation Tailings will not generate acid; Flotation Tailings will have the potential to release constituents of interest Need further explanation or data summarizing what the tailings will produce. PolyMet will add a bentonite amendment to the FTB side slopes, final pond bottom, and final beaches to limit oxidation of sulfide minerals and release of metals. Did PolyMet consider the other extreme of encouraging oxidation to rapidly flush the COC to shorten the length of collection & treatment? Or, filter the water from the tailings and compact them to reduce infiltration of air and water, and then cap with low permeability cover at end of project. This would eliminate the eventual catastrophic failure of the lake releasing the saturated tailings, and reduce or eliminate the long-term water collection and treatment. Document Item #/page 10.1.3.4 LTVSMC Tailings, p 190 10.1.3.5 HydroMet residue, p 190 10.1.3.6 Reagents, p193 10.2 Tailings basin 10.4.4.2 Segregation Appendix 4 Stockpile Drawings General Appendix 4 Stockpile Drawings SKP-012 Appendix 4 Stockpile Drawings SKP-032 Comment Section 10 Mine Waste Characterization Appendix 2, 4, 6, 7 and 17 A summary of the data referenced in Appendix 2 should be placed here Is there a reason why this material is placed as a slurry and not filtered and placed dry with compaction to reduce permeability? If placed dry and compacted, or mixed with a little bentonite and compacted to manage the permeability, the resulting repository would be more stable in the long term. Include a summary of the chemistry data and TCLP data here that justify the conclusions. Summarize the appendix 2 reference here so reader does not need to wade through appendix 2 to find this tidbit. PolyMet's FTB design is intended to achieve multiple objectives, including the following: safe permanent disposal of Flotation Tailings produced over the 20-year LOM The wet closure is temporary, it is not permanent. The geomorphologically unstable slopes will erode back into the pooled water if the slope erosion is not repaired perpetually. The saturated tailings will gush out, oxidize, and cause a variety of downstream water quality problems. After this happens, the State of Minnesota will need to collect the escaped tailings, and rebuild the FTB as a dry closure. This will be very expensive. The last paragraph of 10.2,0 page 194 refers to the next section and needs to be reworded. This section says segregation but the narrative is something else/ The stockpile development drawings by Golder are excellent. They clearly show the stockpile progression and the reason for the location of the piles relative to the wetlands. The use of color made the drawings easier to read than the greyscale drawings provided by Barr. The drawings include more design and construction details that the rules require, but are very informative. The drawing show the final regraded contours of the stockpiles as required in the regulations as well as snap shots of various stages of development. A very minor criticism is that the intermediate stage drawings show the stockpile benches before they are regraded, whereas the permit wording says that the reclamation will be concurrent, so if PolyMet regrades the lower lifts, they will be regraded, the impermeable liner installed, and then vegetated. As a practical matter, this level of detail should only be provided in the annual reports after the mine is in production. More thought was put into the water run off and erosion control of the Cat 1 benches and chutes that into the closure of the FTB. Why? The Cat 1 benches will be sloped towards chutes that convey the water down a 23% slope, whereas nothing is done at the FTB to control erosion of the embankments. However, the overall Cat 1 design is not geomorphologically stable in the long term, so unless the ditches, chutes, and slopes are perpetually maintained there will eventually be erosion that will expose the underlying rock. Document Item #/page Comment Section 10 Mine Waste Characterization Appendix 2, 4, 6, 7 and 17 The design of the Cat 1 cover percolation layer (6-inches” of till) contradicts the logic that the saturated overburden doesn’t take up any volume in the Cat 2/3 pile. Need to know how large the waste rock will be, and the gradation. Are there voids that will fill with till, or not? The size of the rock will also influence the construction of the ditches along the benches and the construction of the down-chutes. This design will be difficult to construct if there are boulders, because the boulders will need to be removed. Appendix 4 Stockpile Drawings SKP-034 Groundwater Containment for Cat 1 GCS series Appendix 4 Stockpile Drawings GCS011 Appendix 6 Pond Seal Pond Seal The plans show the cat 1 groundwater containment system being built in 4 phases as the size of the pile grows. Was any consideration given to segregating the Cat 1 material by %S, so that the clean material could be used for future construction? If this is done, it may change the phasing and the extent of the containment system if the “clean” material was placed at the south end, for example. Will the cutoff and drain be cut into the bedrock as shown? Spectrum has no experience with pond seal. Don Sutton talked to a Geotech engineer who has used it with mixed results. He said that it will be necessary to use more than theoretically estimated because of difficulty evenly placing it. Barr says that a perfect seal is not the goal, they want it to leak to keep the tailings saturated. Document Appendix 2 Waste Characterization Item #/page Comment Section 10 Mine Waste Characterization Appendix 2, 4, 6, 7 and 17 Pond Seal consists of a 1/4 to 3/8 inch diameter rock coated with bentonite. Over time, in a large pond I suspect the roots from aquatic plants would penetrate the seal. I also suspect that if there was a major fluctuation in the water level in the pond that the seal could be washed from the outside edges by wave action, depending on the extent of the riprap beach protection. Is there any correlation between the 5% and the rock type. Is the rock type visually discernable from the cab of the loader? From: Sent: To: Cc: Subject: donsutton@spectrum-eng.com Wednesday, May 24, 2017 4:00 PM Wenz, Zach (DNR); Olson, Michael (DNR); Engstrom, Jennifer N (DNR); Boyle, Jason (DNR); Jordan, Julie E (DNR); Liljegren, Michael W (DNR); Herr, Erika S (DNR); Henderson, Joseph (DNR); Arends, Heather (DNR); Sellner, Jim M (DNR) 'Cecilio Olivier'; 'Stu Grubb'; Woldeab, Irina (DNR); Kunz, Michael (DNR) RE: NorthMet: Appendix review Categories: Important Appendix 2 : Waste characterization. I only did a quick review. I would like to see a detailed plan and commitments  regarding how they intend to track the Cat 1,2,3,4 waste. Including blast hole sampling, dynamic  dispatching, etc. I would like to see a commitment that if they don’t have the waste rock sampled then  can’t be placed in the Cat 1 pile.  They still haven’t identified any of the construction rock sources, so we need to modify the financial assurance to  assume the worst case. They talk about using Cat 1 for construction, but will they? I don’t see any plan  to separately stockpile really clean waste rock for construction or reclamation. My fear is that unless  they use cherty rock from the taconite waste piles, they won’t find any suitable material nearby, and  may want to use something that may generate sulfates. If they construct any roads, pads, etc. with  marginal material, then I think they will need to backfill it into the pit lakes during reclamation. The FA  would need to be adjusted to reflect the cost.  I have a concern with the way they intend to store water on the highwall between the highwall crest and the  overburden. This will encourage water to drain down and oxidize the sulfides in the highwall.    Appendix 4: I reviewed the Golder stockpile drawings. They are acceptable. Some additional Geotech data needs to be  collected before construction. The permit should require that the additional data be collected and the  drawings and plans modified by a P.E., and that a P.E. must supervise the construction and sign (certify)  the as‐built drawings.  Appendix 6: The FTB plan gives me severe indigestion because a lake on top of a pile of sand is inherently unstable, and  irresponsible. The dam embankments are a stair step arrangement that is inherently geomorphologically  unstable, and will erode and cause the ponded water and tailings to escape if it is not maintained and  repaired forever.   I have some questions about the utility of the underdrain on lifts 1 and 5,because they appear to direct run‐off back into  the embankment. If this saturates the embankment, then I am not sure how this affects the  embankment stability.  I have some questions about the FTB groundwater capture. It requires significant pumping, so will require multiple  passive systems at each low point.  Appendix 7: No comment  Appendix 17 Previously commented on this      Donald G. Sutton P.E.  Spectrum Engineering & Environmental  th 1413 4  Ave. North  Billings, MT 59101  Direct: 406‐534‐4660  Mobile: 406‐670‐7270    1 From: Wenz, Zach (DNR) [mailto:zach.wenz@state.mn.us]   Sent: Wednesday, May 24, 2017 9:52 AM  To: Olson, Michael (DNR) ; Engstrom, Jennifer N (DNR) ; Boyle, Jason (DNR) ; Jordan, Julie E (DNR) ; Liljegren, Michael W  (DNR) ; Herr, Erika S (DNR) ; Henderson, Joseph (DNR) ; Arends, Heather (DNR) ; Sellner, Jim M (DNR) ;  donsutton@spectrum‐eng.com  Cc: 'Cecilio Olivier' ; Stu Grubb ; Woldeab, Irina (DNR) ; Kunz, Michael (DNR)   Subject: RE: NorthMet: Appendix review   Importance: High    Hello Section 10 Review Team,    During our last review session we ran out of time to finish going over all the comments and I recall a few of us who had  not yet finished review of appendices: 2, 4, 6, 7, and 17. We are currently working on scheduling a final appendix review  session for section 10 and would like to confirm whether or not you have finished your review to allow for adequate  time to complete prior to meeting.    Please respond to this email indicating whether or not you have reviewed these appendices. We will then set an  appropriate meeting time to finish the review.    Thank You,    Zach    Zach Wenz  Research Scientist   Lands and Minerals Division  Minnesota Department of Natural Resources  500 Lafayette Road  St. Paul, MN 55155‐4045  Phone: 651‐259‐5384  Fax: 651‐296‐5939  Email: zach.wenz@state.mn.us  mndnr.gov            From: Wenz, Zach (DNR)   Sent: Tuesday, April 11, 2017 7:29 AM  To: Olson, Michael (DNR) ; Engstrom, Jennifer N (DNR)  ; Boyle, Jason (DNR) ; Jordan, Julie E (DNR)  ; Liljegren, Michael W (DNR) ; Erika S (DNR)  (erika.herr@state.mn.us) ; Henderson, Joseph (DNR) ;  Arends, Heather (DNR) ; Sellner, Jim M (DNR) ;  'maehl@spectrum‐eng.com'   Cc: 'Cecilio Olivier' ; Stu Grubb ; Woldeab, Irina (DNR)  ; Kunz, Michael (DNR)   Subject: FW: NorthMet: Appendix review   2   Good Morning Section 10 Review Team,    For section 10 of the NorthMet PTM application we have been asked to focus our attention on review of appendices #2,  4, 6, 7, and 17. I will set up a meeting appointment for the first week of May to review our comments. Please continue  your review and add any comments you may have in the appropriate comment spreadsheets. Feel free to contact me  with any questions.    ‐Zach    Zach Wenz  Research Scientist   Lands and Minerals  Minnesota Department of Natural Resources  500 Lafayette Road  St. Paul, MN 55155‐4045  Phone: 651‐259‐5384  Fax: 651‐296‐5939  Email: zach.wenz@state.mn.us  mndnr.gov            From: Woldeab, Irina (DNR)   Sent: Monday, April 10, 2017 4:18 PM  To: Wenz, Zach (DNR)   Cc: Kunz, Michael (DNR)   Subject: NorthMet: Appendix review     Dear Zach,     As previously discussed, following upon our review of PTM sections, we now continue with the review of appendices.  Please see below the appendices that are relevant to the sections of PTM application where you are review team lead.     We realize that the information is a bit confusing, but in general:    At this time please review ONLY the highlighted appendices (others may change for version 2 of the application,  so it makes sense to wait with that review).     Some instances where appendices are mentioned within a section can be very general – if that is the case, we  have probably already asked PolyMet to provide greater specificity. These cases would fall under the “lead‐only”  category listed in the estimated time column, but we can make a note of those instances in the appropriate  appendix comment spreadsheets, found here:  \\2K8FS2\Projects\ALL\NorthMet\In Progress\01 Permit to Mine\Permit Application\PTM comment  spreadsheets    3  For all other instances, the entire team should be involved in the review, much like the section review for the  PTM has taken place. Please contact your team members, use the comment spreadsheets, and feel free to  schedule an internal meeting to discuss comments. I am also happy to schedule those for you and attend/take  notes.     For a full list of the appendices and the times they are mentioned in each section (thank you Mike Olson!) please  see:   \\2K8FS2\Projects\ALL\NorthMet\In Progress\01 Permit to Mine\Permit Application\2017 0215 Appendix refs  in PTM app.pdf    Please let us know if the timing will not work as suggested. Thank you so much in advance!  Irina & Mike    PTM Section  Team Lead  Team Members  Estimated Time  April 24 if lead‐only   Mike Olson     Mike Kunz  May 1 for team   Jennifer Engstrom  involvement   Jason Boyle   Julie Jordan   Mike Liljegren   Erika Herr   Joe Henderson   Heather Arendt   Jim Sellner   Bill Maehl (copy Stu Grubb and  Cecilio Olivier in all  communications)  Appendices: 2 & 17 (mentioned 6 times); 4 & 16.16 (mentioned 8 times); 6 & 16.7 (mentioned 4 times);  7 & 16.8 (mentioned 5 times); 12 (mentioned 12 times); 16 (mentioned twice); 16.5 & 16.22 (mentioned  once); 16.12 (mentioned 11 times); 16.15 (mentioned 10 times); 16.17 (mentioned 7 times)      Irina Woldeab  Planner   Lands and Minerals  10 – Characterization &  Management of Mine  Waste  Zach Wenz  Minnesota Department of Natural Resources  500 Lafayette Road  St. Paul, MN 55155  Phone: 651‐259‐5380  Email: irina.woldeab@state.mn.us   mndnr.gov        4 Spectrum/PolyMet Differences Ver 1.1 April 9, 2017 Type Codes: C=Construction D=Demolition G=General I=Insurance LT=Long Term R=Reclamation WT= Water Treatment PolyMet Potential Resolution Type Spectrum (Don's Comments) C-C1 CS Cat 1 Containment System using generic costs. Cat 1 Containment system is Barr estimate (local contractor price) Use Barr estimate with local contractor price. Barr very familiar with design requirements. Local contractor price includes overhead and reflects local conditions. C-C1 Cvr Cat 1 Cover System using generic costs. Cat 1 Cover is combination of Barr (local contractor price) and SRCE (generic costs) Use complete Barr estimate with local contractor price. Barr very familiar with design requirements. Local contractor price includes overhead and reflects local conditions. C-FTB-B FTB Beach Cover System using generic costs. FTB Beach Cover System is Barr estimate (local contractor price) Use Barr estimate with local contractor price. Barr very familiar with design requirements. Local contractor price includes overhead and reflects local conditions. C-FTB-OF FTB Pond Emergency Overflow using generic costs. FTB Pond Emergency Overflow not included Add Barr estimate with local contractor price. Barr very familiar with design requirements. Local contractor price includes overhead and reflects local conditions. C-FTB-P FTB Pond Cover System using generic costs. FTB Pond Cover System is Barr estimate (local contractor price) Use Barr estimate with local contractor price. Barr very familiar with design requirements. Local contractor price includes overhead and reflects local conditions. D-MSFF Maintenance Service and Fueling Facility demo estimated D-PL Pipeline by location D-PS Used SRCE (Lakehead/Mavo) Demo estimate D-PWR Power Line by location D-RTH Rail Transfer Hopper demo estimated G-DLY Assumes up to a 4 year delay before work can start due to no funds (Minnesota law does not override Federal Bankruptcy Law. I was lead to believe that the rules aren’t enforceable in bankruptcy court. I am not a lawyer.) Assumes no delay before work can start Add a one year delay (holding year) with minimal staff and activity and push all other activity back one year Other states (i.e., Alaska) have adopted this 'holding year' methodology G-DUR Estimate Duration of 212 years and effective NPV Discount Rate 1 to 5% (Term is indefinite until the non-mechanical alternative is demonstrated) Estimate Duration of 50 years and effective NPV Discount Rate about 7% Forest Service Bonding Guidance (2004) states that "100 years can be used as a proxy for perpetuity" if PolyMet's estimte is run out 100 years (rather than 50) very minimal amount is added to the total (130k, or relative % diff of 0.065%) G-FTB Given that there will be impounded water inside the FTB and that if erosion of the embankments is not properly maintained forever, there is a reasonable probability that erosion will head cut back into the pool and release some pooled water, which may wash out some tailings depending on how fast things unravel. If this happens, then it must be cleaned up, and I suspect the FTB closure would need to be rebuilt as some type of geomorphologically stable dry closure. We need to figure out how to address this liability. Perhaps an insurance policy that would be perpetually renewed? For how much? This liability is relatively small in the year 1 case, but I suspect that grows as the height of the embankment and volume and thickness of saturated tailings grows maintenance costs for the basin included in the estimate in perpetuity suggest a PTM permit condition requiring PolyMet to study alternate closure methods for the basin in first few years of operations liability and risk for this is very low in the first few years In Demo price based proposal from local contractor based on specification All pipelines in Demo price proposal from local contractor based on specification Used SRCE (Lakehead/Mavo) Demo estimate All power lines in Demo price proposal from local contractor based on specification In Demo price proposal from local contractor based on specification Use local contractor price Use local contractor price Rationale Local contractor price includes overhead and reflects local conditions. Local contractor price includes overhead and reflects local conditions. Review Lakehead/Mavo to understand what structures included and waste disposal Use local contractor price Use local contractor price Local contractor price includes overhead and reflects local conditions. Local contractor price includes overhead and reflects local conditions. Spectrum/PolyMet Differences Ver 1.1 April 9, 2017 Type Codes: C=Construction D=Demolition G=General I=Insurance LT=Long Term R=Reclamation WT= Water Treatment PolyMet Potential Resolution Type Spectrum (Don's Comments) G-IC need to discuss indirects G-LC Spectrum/EOR questioning whether DNR can get local contractors to provide the same services as the quotes PolyMet has received G-LC indirects table provided as tool for discussion used local contractor quotes; local contractors familiar with MN and site also concern that if DNR has to 'go out to bid' the work, the used local contractor quotes; local contractors familiar local contractors PolyMet has used in their estimate may with MN and site not be chosen in the bidding process DNR determination PTM statutes and regulations do not speak to whether competitive bidding requirements for public contracts would apply - competitive bidding requirements do not apply if the Reclamation services constitute "professional services" and justify use of local service providers; use of local contractors/hiring commitments will be very beneficial for Iron Range DNR determination PTM statutes and regulations do not speak to whether competitive bidding requirements for public contracts would apply - competitive bidding requirements do not apply if the Reclamation services constitute "professional services" and justify use of local service providers; use of local contractors/hiring commitments will be very beneficial for Iron Range G-MU Assumes mark up on long term labor and supplies; this is how Specturn says they are getting paid Assumes no mark up on long term labor and supplies do not include markup G-PW should prevailing wages be used? used local contractor quotes which includes contracting company's wages DNR determination I-ENV Includes Environmental Insurance I-FCY Includes Facility Insurance DNR Project manager: 500k/yr for 4 years then 250k/yr (Need to have DNR tell us what they think. Is the DNR reimbursed for additional admin associated with the work?) LT-DNR LT-DSM $28K/year for dam safety monitoring LT-LFM Landfill Monitoring at $3K/year LT-RDMT Lists road maintenance and snow plowing at $50K/year ($14.5 + $23 = $37.5 OK) LT-RMT Lists maintenance for Cat 1 and FTB covers systems (FTB maintenance, assume $60k/yr. Cat 1 Pile, assumed $15k/yr. May need to increase these and/or add insurance.) Does not include Environmental Insurance in direct costs Does not include Facility Insurance in direct costs SRCE has $503,360 for 4 years for project manager. DNR project manager for 4 years only, then normal DNR costs for reclamation inspections and reporting. Water treatment is PCA $44.6K/year for dam safety monitoring - based on NTS and Barr prices Coal Ash Landfill $3K for 13 years and SW618 Landfill $15.4K for 30 years Row 48 – SRCE includes In Const Mgt Grader at 40 hts/mo for first 4 years (about $70K/yr) and 8 hrs/mo (about $14.5K/yr) for balance and in Other User Snowplowing about $23K for all years Includes allowance (5%) for reseeding all areas seeded but nothing for erosion maintenance on slopes (Pit Walls, Cat 1 Cover, FTB) or keeping water conveyance systems open; plan to use staff for inspection and minor repair and service to provide earthmoving support Rationale indirects do not need to be large percentages; project is well defined, MY1 closure not complex or large; numerous contingencies included in probabilistic water modeling, which is basis for water treatment estimate PolyMet not a Superfund site, so not comparable to Spectrum's work PTM statutes and regulations do not speak to whether prevailing wages apply; prevailings wages try to account for future uncertainty by assuming the highest cost; however, annual updates are made to the FA estimates - therefore, there is no need to addresss future uncertainties in wage costs by overestimating those costs on the front end discuss and agree; DNR determination can be added as indirect cost discuss and agree; DNR determination can be added as indirect cost DNR determination Site not complex or large in a MY 1 closure scenario; site becomes stable and work becomes routine after first few years use PolyMet local firms doing work now use PolyMet estimates from firms doing the work now use PolyMet need to include erosion maintenace on slopes and keeping water conveyances open; proposed costs for tailings basin maintenance provided to Spectrum via email; need to discuss and agree suggested costs for tailings basin maintenance based on what is going on now at tailings basin Spectrum/PolyMet Differences Ver 1.1 April 9, 2017 Type Codes: C=Construction D=Demolition G=General I=Insurance LT=Long Term R=Reclamation WT= Water Treatment Spectrum (Don's Comments) Type PolyMet Potential Resolution $48K/yr - milage x $/mile for vehicles (Difficult to estimate, $21.4K/yr for first 4 years then $10.7K for vehicles that’s why I use $1/mile and miles. Also, if on call during Need to include insurance, liscense, fuel and routine LT-VCL purchase assume 7 year live - insurance, liscense, fuel night, then, then need to drive company trucks home. So, maint and routine maint not included mileage piles up.) Adds supplies, sampling labor, pickup, markup to water quality monitoring lab analysis price - frequency monthly to Does not add supplies (included in Pace price), sampling quarterly when reclamation done (I suggest we get labor (by site Labor), pickup (use site vehicle), markup together and iron out the costs after the sampling details LT-WQM to water quality monitoring lab analysis price use PolyMet are known. frequency from monthly to quarterly after 5 years I was assuming a 3rd party contractor would perform the (NPDES mod for closed facility starts new 5 year cycle) work. It would be easier for the State to manage, and the quality would be better.) HRF Disturbed Area Reclamation is Barr estimate (local R-HRF HRF Disturbed Area Reclamation not included Use Barr estimate with local contractor price. contractor price) R-OSLA R-Road R-S&P Overburden Storage and Laydown Area assumed to contain Overburden Storage and Laydown Area assumed to to 4.7M CY that needs to be consolidated into one pile. be mostly empty with only minor grading. Roads by location All roads in Demo price based proposal from local contractor based on specification About half cost of Spectrum - liner and backfill -(SRCE Pond dimensions have been obtained from Large Figure Sumps and Ponds - sediment and liner disposal and backfill 4 of Dimensions from the to grade - about twice cost of PolyMet "Water_Management_Plan_-_Mine." The pond depths have been approximated from Table 4-1 of the same) Rationale see Pace price list Barr very familiar with design requirements. Local contractor price includes overhead and reflects local conditions. use PolyMet likely to be empty Use local contractor price - confirm method for haul roads Local contractor price includes overhead and reflects local conditions. discuss and agree; DNR determination depressions remain to promote the formation of wetlands Use $315/acre $315/A but looks close to D&T (local contractor) price of $295 + sales tax R-Seed Reclamation seeding Used $548.48/acre SRCE uses $315/acre R-SP Includes Performance Bond for all activities No Performance Bond Included Add 1% to Barr Estimates for Cat 1 Stockpile Cover and Other reclamation activities such as temporary stockpile relocation Cont System Extension and FTB Beach/Pond Bentonite and footprint reclamation, pond reclamation, etc. are straight Ammendment and Lakehead Demo - Add 1.75% to forward earthmoving and do not require performance bonds. Mavo Abatement - See Const Addons Tab No Final Engineering included Final engineering will be completed early in the clsoure term - no Add Final Engineering ($389K) into Barr Construction need for redesign in MY1 clsoure as reclamation area is small and Estimates for Cat 1 Stockpile Cover and Cont System site will stabilize in a short period of time; Other reclamation Extension and FTB Beach/Pond Bentonite activities such as temporary stockpile relocation and footprint Ammendment - See Final Design Engineeering Memo reclamation, pond reclamation, etc. are straight forward earthmoving and do not require engineering designs. R-SP Includes Final Engineering at $50K/year R-SP Includes Mobilization for all activities R-SP Cat 2/3 Stockpile Haul Distance 8,740ft and Unit Cost of $1.83/T Mobilization included in Cat 1 Stockpile Cover and Cont Add Mobililization to Earthmoving components of System Extension, FTB Beach/Pond Bentonite SRCE; This would cover all portions where equipment Ammendment, Demo and Abatment requiring mobilization would be required Cat 2/3 Stockpile Haul Distance 7,554ft and Unit Cost of $1.38/T PolyMet provided a table that showed the breakdown of contractor quotes that included mobilization vs costs that did not (i.e., SRCE earthworking components) Have Barr add haul from center of stockpile to center COT memo is source for most dimentions that change over time of backfill in COT memo. Discuss derivation of unit Barr has pit shells and stockpile elevations costs. Spectrum/PolyMet Differences Ver 1.1 April 9, 2017 Type Codes: C=Construction D=Demolition G=General I=Insurance LT=Long Term R=Reclamation WT= Water Treatment Spectrum (Don's Comments) Type PolyMet Potential Resolution Rationale Cat 2/3 Liner (membrane and soil components) and Cat 2/3 Liner (membrane and soil components) and COT memo has liner acres and drain pipe footage for all years. underdrain piping based on local contractor price R-SP underdrain piping based on estimates of quantity (CY or Use local contractor price per acre or foot Local contractor price includes overhead and reflects local estimates per acre for liner (membrane and soil) and tons) to be moved conditions. per foot of underdrain pipe FEIS states in Table 5.2.4-2 that proposed reclamation for cat 2/3 stockpile is wetlands vegetation. 6" of soil would be optimum Cat 2/3 Footprint Reclamation includes replacing 18" OVB Cat 2/3 Footprint Reclamation includes replacing 6" R-SP Use 6" amount of soil to establish/restablish wetland vegetation; also, and soil OVB and soil Revegetation SOP Reviewed and Accepted by agency for FEIS stated 6" Have Barr add haul from center of stockpile to center Cat 4 Stockpile Haul Distance 4,055ft and Unit Cost of Cat 4 Stockpile Haul Distance 4,574ft and Unit Cost of COT memo is source for most dimentions that change over time R-SP of backfill in COT memo. Discuss derivation of unit $1.57/T $1.24/T Barr has pit shells and stockpile elevations costs. Cat 4 Liner (membrane and soil components) and Cat 4 Liner (membrane and soil components) and COT memo has liner acres and drain pipe footage for all years. underdrain piping based on local contractor price R-SP underdrain piping based on estimates of quanity (CY or Use local contractor price per acre or foot Local contractor price includes overhead and reflects local estimates per acre for liner (membrane and soil) and tons) to be moved conditions. per foot of underdrain pipe FEIS states in Table 5.2.4-2 that proposed reclamation for cat 4 stockpile is wetlands vegetation. 6" of soil would be optimum Cat 4 Footprint Reclamation includes replacing 18" OVB Cat 4 Footprint Reclamation includes replacing 6" OVB R-SP Use 6" amount of soil to establish/restablish wetland vegetation; also, and soil and soil Revegetation SOP Reviewed and Accepted by agnecies for FEIS stated 6" Have Barr add haul from center of stockpile to center OSP 970,000 tons - Haul Distance 8,410ft and Unit Cost of OSP 2,275,000 - Haul Distance 9,052ft and Unit Cost of COT memo is source for most dimensions that change over time R-SP of backfill in COT memo. Discuss derivation of unit $1.81/T $1.51/T Barr has pit shells and stockpile elevations costs and tons. OSP Liner (membrane and soil components) and OSP Liner (membrane and soil components) and COT memo has liner acres and drain pipe footage for all years. underdrain piping based on local contractor price R-SP underdrain piping based on estimates of quanity (CY or Use local contractor price per acre or foot Local contractor price includes overhead and reflects local estimates per acre for liner (membrane and soil) and tons) to be moved conditions. per foot of underdrain pipe FEIS states in Table 5.2.4-2 that proposed reclamation for OSP stockpile is wetlands vegetation. 6" of soil would be optimum OSP Footprint Reclamation includes replacing 18" OVB and OSP Footprint Reclamation includes replacing 6" OVB R-SP Use 6" amount of soil to establish/restablish wetland vegetation; also, soil and soil Revegetation SOP Reviewed and Accepted by agnecies for FEIS stated 6" If Barr cannot find support for PolyMet assumption, .5 FTE for Elect/Mech (We operate Landusky a floater Assumed to be included in facility replacement (repair Prefer to use service for specialized work that cannot be done by WT-MNT add a percent of OPEX for supplies and craft services elec/mech) or replace) staff as needed .05% OPEX for materials and supplies (Row 9, Opex is supplies only. Major replacement parts (membranes) in OPEX. Rest If Barr cannot find support for PolyMet assumption, WT-MNT Row 13, is major component replacement assumed to be included in facility replacement (repair add a percent of OPEX for supplies and craft services Row 25, is a category for misc odds’n ends. Gaskets, or replace) as needed grease, oil, tools, etc.) Mine Site Annual Capital Replacement Allowance (zero) (I Row 41 – the equipment in service is CPS and TWP doubt if anything will be replaced in 1st year. $5,855 isn’t which will be in long term service once pit water level WT-MNT much. use PolyMet reaches overflow elevation - $5855 annual replacement If mine closes, this will stop when plant shuts down and is cost – looks like Barr memo has this for 42 years removed) Spectrum/PolyMet Differences Ver 1.1 April 9, 2017 Type Codes: C=Construction D=Demolition G=General I=Insurance LT=Long Term R=Reclamation WT= Water Treatment Spectrum (Don's Comments) Type PolyMet Potential Resolution Rationale $20K/yr for sampling and analysis for operations (I didn’t see any money to handle the internal water chemistry Assumed none - influent and effluent streams included WT-MON monitoring. There needs to be hourly or daily checking in overall water monitoring and internal streams Discuss and agree doesn’t there? What about the cost to calibrate and monitored with instrumentation maintain any probes?) Includes $1.6M for NMT development for FTB but DNR has stated that cost for development (not construction or WT-NMT Includes $300K for NMT development Use PolyMet and add West Pit and Cat 1 nothing for Cat 1 or Pit Overflow operation) of NMT be included From: Sent: To: Cc: Subject: Dostert, Dana M (DNR) Thursday, June 01, 2017 1:34 PM Cecilio Olivier; donsutton@spectrum-eng.com; Boyle, Jason (DNR) Kunz, Michael (DNR); Stu Grubb RE: PolyMet Tailings Dam Comments Appendix 6 Slope stability models were run on the perimeter embankments using a variety of scenarios.  Those scenarios included  both drained and undrained embankment conditions.  In addition, slope stability models were computed with water  levels within the basin at levels expected from the 24‐Hour Probable Maximum Flood.  These models were run at  numerous lifts all the way to closure.  The specific information can be found in the geotechnical data package.    The geomorphological issues are essentially why I favor dry closure.  I would like to be able to see Polymet be able to  walk away from a site that no longer has seepage issues and can be allowed to revert to forest.  Wet closure will not  allow that.    Dana D.    Dana Dostert PE, PG  Senior Engineer – Dam Safety  MN Department of Natural Resources  500 Lafayette Road, St. Paul, MN 55155‐4032  (651)‐259‐5663    mailto:dana.dostert@state.mn.us      From: Cecilio Olivier [mailto:colivier@eorinc.com]   Sent: Thursday, June 01, 2017 12:01 PM  To: Dostert, Dana M (DNR) ; donsutton@spectrum‐eng.com; Boyle, Jason (DNR)    Cc: Kunz, Michael (DNR) ; Stu Grubb   Subject: RE: PolyMet Tailings Dam Comments Appendix 6  Importance: High    Following Don’s and Dana’s comments below, it seems to me that there a couple of issues we need to address:    1. The tailings dam could be geotechnically stable but geomorphologically unstable. Dana, Jason – do you know if  the geotechnical stability analysis was performed under spillway overflow AND saturated embankment? If not,  do you recall the rational for not doing it this way?  2. All the points raised by Don regarding geomorphological issues are important and, short of a dry closure, they  will require significant perpetual maintenance. Right now, the cost tha PolyMet is assuming for this maintenance  tasks appears to be about $100K (this needs verification but seems very low to me). I know it is difficult, but we  really need to estimate these annual costs for long term FA calculation purposes. Don, could you do that?    My opinion is that, at a minimum, we need to include:   Annual maintenance costs for regular erosion patching, tree/vegetation replacement, regrading, etc.   Annualized cost of perpetual operation, maintenance, and capital replacement of the pumping system  to maintain appropriate tailings water levels (pumping in or pumping out)  1  Annualized cost of major capital improvements like bentonite re‐application (say every 10‐20 years),  structural issues, etc.  Annual monitoring and 3rd party dam safety consultant     It is not going to be cheap….. we need to know sooner than later.    Thanks,      Cecilio Olivier, PE EOR: water ecology community d: 651.203.6001 o: 651.770.8448 EOR - CELEBRATING 20 YEARS!    From: Dostert, Dana M (DNR) [mailto:dana.dostert@state.mn.us] Sent: Thursday, June 1, 2017 8:17 AM To: donsutton@spectrum-eng.com; Cecilio Olivier; 'Steve Gale'; Kunz, Michael (DNR); 'Nate Lichty'; Stu Grubb Cc: Boyle, Jason (DNR) Subject: RE: PolyMet Tailings Dam Comments Appendix 6   Hi Everybody,    Just as an update, I talked to Tom Radue earlier this week about the under drain.  In the most recent plans revised about  two weeks ago, they have renamed the two underdrain layers as foundation layers.  Their primary purpose will be to  provide foundation support for the perimeter dam and they will be constructed of gravel and coarse tailings.   They were  also meant to collect seepage waters and draw them back into the stack.  The drainage area for these seepage collection  sites is quite small, being the outside face of the dam, so there is unlikely to be a large volume of precipitation entering  the stack.  There groundwater model has not shown any issues with rising groundwater levels at this time.     We have not yet received the redesigned plans.  More discussion will be needed.    Thanks,,    Dana D.      Dana Dostert PE, PG  Senior Engineer – Dam Safety  MN Department of Natural Resources  500 Lafayette Road, St. Paul, MN 55155‐4032  (651)‐259‐5663    mailto:dana.dostert@state.mn.us      From: donsutton@spectrum‐eng.com [mailto:donsutton@spectrum‐eng.com]   Sent: Wednesday, May 31, 2017 5:18 PM  To: 'Cecilio Olivier' ; 'Steve Gale' ; Kunz, Michael (DNR)  ; 'Nate Lichty' ; 'Stu Grubb'   Cc: Boyle, Jason (DNR) ; Dostert, Dana M (DNR)   Subject: PolyMet Tailings Dam Comments Appendix 6    2 Steve Gale, Nat Lichty and I discussed the FTB drawings in Appendix 6 today.    1. As shown on the drawings, (FTB‐009) none of us completely understand the function of the 4‐foot thick coarse  LTV tailings underdrain beneath lifts 1 and 5 along the north face of the embankment.  It appears that the  underdrain will collect and direct surface run‐off into the embankment, potentially saturating the embankment,  and reducing the stability. This item requires further explanation and review. (see  first image below) We don’t  think that the purpose of the underdrain is to collect surface water, but that’s what is illustrated.     2. I wasn’t tasked to look at the dam stability, so don’t know if the stability analysis assumed the embankment was  saturated or dry.  The analysis needs to be done considering rainy wet conditions where the surface run‐off is  being forced into the embankments.  This will shift the phreatic surface closer to the embankment, especially  below lift 5. Below, I describe some other scenarios where erosion can alter the phreatic surface assumptions  and potentially cause embankment failure. The analysis also needs to be made when the pond is overflowing  the spillway and the embankment is saturated. This is the worst case, unless an earth quake occurs at the same  time.    3. The stair step FTB embankment sealed with bentonite is geomorphologically unstable and will erode, potentially  cutting back into the pooled water, releasing the water and saturated tailings. Initially, surface water will collect  in the horizontal ditch/ponds along the toes of lifts 1 and 5, and infiltrate into the embankment via the  underdrain and the coarse LTV tailings beneath lift 1. Later, after the bentonite soil erodes from the slopes, the  ditches will fill, plugging the underdrain, forcing the water to overflow the bench and cause head cutting in the  non‐cohesive tailings. If the FTB is to remain as a permanent structure without perpetual maintenance, then I  recommend that the embankments be designed using established geomorphologic land reclamation principals.  Otherwise there is a high probability that the embankments will eventually fail due to erosion, and  catastrophically release the saturated tailings.    4. As illustrated in Drawing FTB‐024, the portion of beach protected by riprap appears to be too narrow, but the  width is subject to change.  The total extent of the riprap needs to be designed as part of the closure.  The size  and thickness of riprap need to be justified based on wave action and ice. If the water level fluctuates, wave  action could erode above or below the riprap, thus setting up a head cutting scenario from the inside towards  the embankment.  This could lead to piping and embankment failure. (see second image below).  The 625 foot  beach slopes 1%, then transitions to a 3% slope to the pond bottom.  If the water level drops below the  elevation predicted, then the 1% to 3% transition nick point could initiate a head cut that will run back to the  embankment.  This could trigger piping by allowing water a clear path into the coarse embankment fill when the  water level rises.  This may not be an issue if the site is perpetually managed and repaired, but will be an issue if  the site is abandoned. The water level will not remain constant unless it is managed.  If it is not managed, then  depending on the bentonite efficacity, the pond could either periodically dry up or over fill.  Climate change  makes precipitation predictions 100 or 200 years from now impossible, so the design needs to assume the worst  case.   The range of water level possibilities needs to be addressed in the closure design. If the water level drops  lower than designed due to higher infiltration rates or lower precipitation, then the geochemistry assumptions  will change as the tailings dry out and oxidize.    The design of the FTB in this permit application will require perpetual maintenance to ensure it will not fail and release  the tailings.  Placing bentonite on the embankment and interior surfaces will increase the run‐off and the erosion  rate.  The stair‐step design is geomorphologically unstable. The methods and assumptions used to place the bentonite to  control the infiltration and tailings saturation are unsubstantiated, and wishful thinking. We do not believe it will  function as intended, because of the unproven application methods.    3 1m- 1m urface entdnite amended reduees ?ltratien . ncreasing run-eff nd erdsien Surface water ceilectien dih directs flew into underdrain mass Tmuss t? Surface water cellectien ditch directs flew into underdrain EMTEFI-UR FACE er mu lElearse ailiings ?1+m       Donald G. Sutton P.E.  Spectrum Engineering & Environmental  th 1413 4  Ave. North  Billings, MT 59101  Direct: 406‐534‐4660  Mobile: 406‐670‐7270    5