Royal Bank of Canada Europe Limited INDUSTRY | … notes/2010...Should this be the case, we would expect a capital raise in excess of US$100 million. Jubilee Platinum (“Jubilee”)

INDUSTRY | COMMENTFEBRUARY 19, 2010

PGMs & South African Golds

Trip Notes - Who Needs What When

Event

RBC visited several South African PGM and gold projects around the IndabaConference. A common theme we discerned was a need to raise funds forprojects; on pages 3-4, we summarise our estimates of each company'srequirements.

• Kroondal vs Crocodile River: Successful Contrasts - AquariusPlatinum's Kroondal and Eastern Platinum's Crocodile River are shallowUG2 mines that are doing quite well relative to many other PGMoperations. Yet, the two are virtually chalk-and-cheese operationally. AtKroondal, Aquarius has taken an innovative approach, mechanically-mininga wide cut and using a Dense Media Separator. Heavily-faulted CrocodileRiver is much steeper-dipping, requiring conventional mining. In our view,good management of dilution, ground conditions and development havebeen the keys to success here, achieved in the face of a host of detractorsand nay-sayers.

• Fiery Innovation At Jubilee & Sylvania - Jubilee's ConRoast smeltingtechnology could potentially revolutionise the PGM industry. Intriguingly,the company is investigating a still-secret "third" downstream refiningtechnology that could apparently see much higher payabilities than themore-conventional converting and leaching routes. Jubilee and Sylvaniarecently agreed to pool their smelting and tailings retreatment expertise toexploit resources previously deemed uneconomic (typically low gradeand/or high chrome).

• Consolidation Opportunities Abound Around The Pilanesberg -Rooderand could potentially be much bigger for Platinum Australia thancurrent flagship Smokey Hills. To us, consolidation of the variousproperties in the northern Pilanesberg area would make economic sense;myriad deal combinations seem possible, if the parties can agree terms.

• New Gold Kids Modder East & Burnstone vs South Deep-zilla - GoldOne's Modder East is a new gold mine in ramp-up; the current mine life isshort, but there are potential life-extending resources to be explored on theproperty and around it. Great Basin's larger Burnstone project, soon tocommence production, is trialling longhole stooping. This, if successful,could yield a range of potential benefits. These mines are relatively-small,shallow (<500m), low-risk operations compared to Gold Field'schallenging, mammoth South Deep. The mine is targeting 300 koz in FY10(almost as much as Modder East and Burnstone at steady-state combined)and >750 koz by FY14. Unusually, South Deep will be applyingmechanised longhole and drift-and-fill methods at depths of 2.5-3 km(!).

• The Golden "Oldies" - Last but not least, we also visited DRDGOLD'sErgoGold, Central Rand Gold's CMR and Pan African's Barberton.ErgoGold and CMR are mining "old" Witswaterand gold resources, exceptthat ErgoGold is a large-scale low-risk surface tailings retreatmentoperation whereas CMR is trialling innovative mining and processingtechniques virtually under Johannesburg. Barberton, meanwhile, is an oldgreenstone belt mine that just keeps on going.

Priced as of prior trading day's market close, EST (unless otherwise noted).All values in USD unless otherwise noted.

Royal Bank of Canada Europe Limited

Leon Esterhuizen (Analyst)(+44) 207 653-4154;[email protected]

Arnold van Graan (Associate)(+44) 207 653-4639;[email protected]

Yuen Low (Associate)(+44) 207 653-4647; [email protected]

For Required Non-U.S. Analyst and Conflicts Disclosures, see page 39.

2

Companies Visited Aquarius Platinum .....................................................................................................................................................................................5 Eastern Platinum........................................................................................................................................................................................7 Jubilee Platinum.......................................................................................................................................................................................12 Platinum Australia ...................................................................................................................................................................................16 Platmin.....................................................................................................................................................................................................18 Sylvania Resources ..................................................................................................................................................................................19 Central Rand Gold ...................................................................................................................................................................................21 DRDGOLD..............................................................................................................................................................................................24 Gold Fields...............................................................................................................................................................................................27 Gold One International ............................................................................................................................................................................29 Great Basin Gold .....................................................................................................................................................................................33 Pan African Resources.............................................................................................................................................................................35

Mines & Projects Visited Kroondal PGM Mine .................................................................................................................................................................................5 Crocodile River Mine ................................................................................................................................................................................7 Mintek 3.2 MV PGM Smelter..................................................................................................................................................................12 Rooderand PGM Project ..........................................................................................................................................................................16 Pilanesberg Platinum Mine ......................................................................................................................................................................18 Samancor Dump Operations ....................................................................................................................................................................19 CMR Trial Mine ......................................................................................................................................................................................21 ErgoGold..................................................................................................................................................................................................24 South Deep Gold Mine ............................................................................................................................................................................27 Modder East Gold Mine...........................................................................................................................................................................29 Burnstone Gold Mine...............................................................................................................................................................................33 Barberton Gold Mine ...............................................................................................................................................................................35

PGMs & South African GoldsFebruary 19, 2010

3

Who Needs What When One of the conclusions we drew from these visits is that many companies will need to raise funding for their projects. The situations of these companies are summarised below.

Aquarius Platinum (“Aquarius”)

• Funding Not Required - We do not currently anticipate Aquarius requiring any external funding unless a significant acquisition is made.

Eastern Platinum (“Eastplats”)

Eastplats currently has around C$30 million in working capital.

• Internal Funding Possible - Crocette, Spitzkop and Mareesburg could be funded from internal cash flows, according to our model.

• Favourable Environment Could Tempt Capital Raising - However, given the rise in Eastplats’ share price over recent months, we believe Eastplats could well be tempted to come to the market to speed up Spitzkop and/or Mareesburg. Should this be the case, we would expect a capital raise in excess of US$100 million.

Jubilee Platinum (“Jubilee”)

• Enlarged Scope Requires Funding - Jubilee currently has around R240 million in the bank, mostly raised in 2009 for a “simpler” Middelburg PGM toll smelter project. The project’s scope has grown, however, and this means that additional funds will be required.

• Partner Could Contribute - We understand that Jubilee will either get its “Middelburg partner” (unspecified major, which will be providing most or all of the feed) to contribute in return for some sort of interest or stake.

• Market Alternative – The alternative would be to come to the markets to raise the requisite funds, probably around mid CY10. In our view, Jubilee would look to raise at least R400 million (equity and/or debt), but we see a figure of R500 million or more as more likely, as Jubilee has other projects to fund.

• More Furnaces Further Out – Jubilee will have to fund 70% of the capex for the SmeltCo joint venture with Sylvania. For the moment, we estimate that Jubilee’s share of plant establishment costs (including one furnace) would be in the region of US$70 million, and that this funding would have to be raised around mid CY11.

Platinum Australia (“PLA”) At the end of Q2-FY10, PLA had just under A$23 million of cash and was projecting an A$19.2 million cash balance for Q3-FY10.

• Small Raise Possible In CY10 - If all goes to plan at Kalplats, assuming Smokey Hills is sufficiently cash-positive and assuming equity funding of 40% or less, PLA should not need to come to the market in CY10. However, prudence would suggest a small capital raise (say US$5-10 million). A larger raise would likely be required if the project scope is significantly enlarged.

• Rooderand Raising in CY11, Possibly Even CY10 - PLA could come to the market to raise capital for the construction of Rooderand in late FY11, or if there is sufficient investor appetite, possibly even in CY10. We believe the project capex is likely to be around R0.8-1.0 billion.

• PLA Might Raise 100% Of Rooderand Funding - PLA is earning into 65% on feasibility study completion, but could earn a further 5% by arranging financing for the project’s completion, which we interpret as meaning that PLA will raise and “lend” the black empowerment partner 30% of the project capex.

Platmin Platmin is planning to raise capital, but we believe for acquisitions rather than for the Pilanesberg operation.

• Sedibelo & Magazynskraal - We believe that that the Pallinghurst Consortium is trying to acquire Sedibelo, and there is also a Bankable Feasibility Study (“BFS”) underway at Magazynskraal (jointly-owned by the Pallinghurst Consortium and the Bakgatla tribe).

• Vending Into Platmin Makes Sense - In our view, it would make sense for Platmin to raise funds externally to “purchase” these projects from the Pallinghurst Consortium: there should be operational synergies, plus the issue of equity to external parties would improve the liquidity of Platmin’s shares.

• Many Variables To Funding Requirements – Platmin already has three significant non-Pilanesberg projects on backburner: Grootboom (which we believe is essentially ready to go), Mphahlele and Loskop. Platmin could choose to develop them, or to dispose of one or more to reduce its funding requirements.

The amount(s) to be raised would therefore likely to be dependent on a range of variables: the purchase prices, scale of project(s), whether Platmin decides to build a smelter at Pilanesberg and the proportions of debt and equity to be issued. We expect that the total cost to Platmin will be at least several billion rand.

Sylvania Resources (“Sylvania”)

• Grass Valley: A Lot Depends On Timeline - Assuming construction of Grass Valley commences on schedule (H1-FY13), we anticipate Sylvania having to raise funds in FY12. Exactly how much will depend on how much cash the SDO operations amass during that time, and the pace of 100 ktpm plant and 10 MW DC furnace rollouts. Disregarding any cash on the balance sheet, we estimate that Sylvania would need to raise around R300-400 million to get the project started, including one 100 ktpm plant – we assume for now that the remaining plants can be funded out of cash flow. Sylvania would also have to fund 30% of the cost of the DC furnaces; for the moment, we estimate that Sylvania’s share of plant establishment costs (including one furnace) would be in the region of US$30 million.

• Everest North: Internal Funding Could Be Used For Grass Valley? – We currently estimate capex at just under R400 million, funded entirely from SDO cash flows, with construction starting in late FY11.

Clearly, if Everest North doesn’t go ahead, or is delayed significantly, external funding requirements for Grass Valley could be a lot lower. Of course, if both projects go ahead, Sylvania could opt to fund Grass Valley (mostly, on our numbers) internally and Everest North externally.


4

Central Rand Gold (“CRG”) Trial mining is on-track for a revised Competent Person’s Report (“CPR”) to be issued in Q2-CY10 (March-April).

• Fund Raising After CPR - A fund raising will then follow, currently envisaged at CRG to be in the vicinity of US$25-35 million. This includes US$10 million for a pump station but does NOT include an optical sorter or CIL upgrade.

• Larger Raise Possible - Given trial stoping results to date, should trial processing prove successful, we would be unsurprised to see the fund-raising figure rise to around US$45-50 million.

DRDGOLD (“DRD”) DRD is investigating potential synergies between Ergo and DRD’s Crown operations. A R290 million pipeline linking the two is being mooted; total project capex would be R345 million.

• Aurora Sales Could Raise Significant Cash - This project was one of the motivating factors behind the sales of 60% of Blyvoor to Aurora for R296 million, and of the ERPM plant for R20 million.

• DRD To Study Aurora Tailings - Aurora has paid R5 million towards the ERPM plant purchase and agreed to allow DRD to conduct a feasibility study on Aurora’s Grootvlei and Marievale tailings dams near Ergo.

• Cash Or Option? - DRD may choose to receive the outstanding R15 million in cash; alternatively, DRD could take an exclusive 10-year option to retreat those tailings dams, which would require capex spend of R260 million.

• Option Exercise Likely To Trigger Fund Raising - Should DRD exercise the Aurora tailings option, we believe that DRD debt or equity financing will be required, given the company’s December 2009 cash balance of R162 million.

• Ergo Phase 2 Would Also Require External Funding - We believe that capex for Ergo’s Phase 2, which is at feasibility study stage, would be in the order of several billion rand. Therefore, DRD would almost certainly have to raise equity and/or debt financing should this project be given the go-ahead.

Gold Fields • Funding Not Required - We do not currently anticipate Gold Fields requiring any external funding unless the company makes a significant acquisition.

Gold One International (“Gold One”)

• Standby Debt Facility Being Arranged For Convertibles - Gold One’s convertible bondholders have a put option exercisable on December 12th, 2010. Although GDO was already cash-flow positive in December 2009 (including capital - but not bond interest), we do not expect GDO to have generated sufficient cash in the unlikely event that all (or a substantial portion) of bondholders elect to exercise. Prudently, GDO is looking to arrange a debt facility to cover any funding requirements that should arise as a result, which could be up to around US$60-64 million (depending on bond repurchases in Q1-CY10).

• Ventersburg – The Bankable Feasibility Study is currently expected in 2011. Ventersburg capex should cost around US$200-300 million, some of which we expect will be funded by Gold Fields. We estimate that Gold One would still have to raise around US$100-150 million (or more, depending on Area 1’s proportional contribution to the overall project) in 2011. This amount could be reduced somewhat via the spin-out or sale of Vlakfontein.

Great Basin Gold (“Great Basin”)

• Funding Not Required - Great Basin should not require additional funds unless there are significant project delays or the gold price falls significantly.

Pan African Resources (“Pan African”)

Pan African currently has around £4.2 million of cash on its balance sheet.

• Barberton Cash Cow – Barberton is essentially a self-funding cash cow that throws of excess cash (after capex) of around £4-5 million annually, depending on the gold price.

• Phoenix Could Be Funded Internally - The latest capex estimate is R100 million, or £8.5 million at R11.8/£. Additionally, Pan African has agreed with International Ferro Metals (“IFM”) a fixed “price” of R80 million (£6.8 million) for the Phoenix plant site and acquisition of IFM’s 25% Net Profit Interest in the PGMs from tailings arising from IFM’s Lesedi property. Depending on timing, it may be possible for Phoenix to be funded out of Barberton’s excess cash flows, but a small capital raise (£5-10 million) in H2-CY10 seems to us more likely.

• Manica Too – Assuming heap leaching testwork is successful and sufficient oxide resources can be obtained, a feasibility study could be conducted in FY11. If construction commences in FY12 or later, Pan African’s share of capex could potentially be funded by cash flows from Barberton and Phoenix.

A potential complicating factor is the level of dividends Pan African decides to pay (if any) – dividends on the scale of that paid in FY09 (£2.8 million) would wipe out most of Barberton’s free cash flow


5

Aquarius Platinum

Kroondal PGM Mine RBC visited the Bambanani Shaft (formerly 3 Shaft) at Aquarius Platinum’s (LSE & ASX: AQP; “Aquarius”) Kroondal mine on Thursday, February 4th.

About Kroondal Kroondal, located around 120 km northwest of Johannesburg, is a 50:50 joint venture with Anglo Platinum. It comprises: • Two blocks of ground, Kroondal and Townlands, both immediately up-dip of Anglo Platinum’s Rustenburg. • Four decline sections: Kopaneng (previously: Central), Simunye (East), Bambanani (No. 3) on the Kroondal block, and K5 at

Townlands. Plans for a second shaft at Townlands, K6, were put on ice during the global downturn. • Two concentrator plants (K1 and K2) with a combined capacity of 570 ktpm; there is a life-of-mine offtake with Anglo Platinum.

In FY09, Kroondal produced 422 koz PGMs, of which 211 koz was attributable to Aquarius – representing just under half of Aquarius’ overall attributable production.

Mechanised Bulk Mining And DMS Processing - Keys To Kroondal’s Low Costs Aquarius mines the UG2 reef at Kroondal Mine. On the property, the UG2 chromitite reef (“Main Seam) is typically around 0.8m thick and grades around 6 g/t (comparable to Anglo Platinum’s neighbouring Rustenburg property). Above, there is a chromitite Leader Seam (or “Triplets”, around 30cm-thick, grading about 3.3 g/t) separated from the Main Seam by an essentially-barren waste pyroxenite parting approximately 60cm thick. • On many similar properties, only the UG2 reef is mined, usually by conventional mining methods. This results in significant

dilution and high production costs. For example, at Rustenburg, the UG2 reef is typically 0.6m-0.8m thick, whereas the mining cut is around 0.95m-1.05m, resulting in a reserve grade of around 3.6 g/t.

• At Kroondal, the Leader Seam, Parting and Main Seam are all taken as a single package (1.8m average stoping width), via mechanised bord-and-pillar mining (facilitated by a relatively-gentle dip of around 9°). This results in a lower mining cost, but also a lower head grade of around 2.7 g/t - which would misleadingly tend to suggest high processing costs.

• In actuality, the Run-of-Mine (“ROM”) material is scalped underground (removing waste rock representing around 8% of mass) before being conveyed to a surface Dense Media Separator (“DMS”) plant, which removes still further waste (around 30%-35% of mass). Consequently, very little waste actually goes into the plant, resulting in low plant operating costs.

• Kroondal’s costs are also helped by the fact that mining is shallow, to a maximum depth of around 400m-450m at present.

Downdip Expansion Potential Across The Borders • Intriguingly, all three recently-closed Rustenburg shafts are adjacent to Aquarius’ mining areas. Notably, Bleskop is downdip of

Kroondal and Brakspruit downdip of Marikana. • We believe there is potential for Aquarius to extend decline-based mechanised mining at Kroondal beyond 2018 by crossing the

mine’s boundary into unmined ground on the western portion of Anglo Platinum’s Bleskop shaft. • The deeper depth is likely to raise costs somewhat, but we believe Aquarius’ methodology should allow the company to maintain

costs far below that which was incurred by Anglo Platinum.


6

Exhibit 1: Bord-and-Pillar Mining at Kroondal. Note the width of the reef, parting and stopes (high enough to stand comfortably in!), the cleaning by LHD and the initial underground scalping of waste via grizzly.

Source: RBC Capital Markets

Blue Ridge - Full Cut Being Taken • Following a fall of ground incident in December 2009, Aquarius has decided to mine the full reef at Blue Ridge, rather than taking

just the efficient cut. • Given current metal prices, we believe this not to be a major problem (efficient cut is preferable at times of lower metal prices). • Blue Ridge is expected to reach steady-state by the end of FY10. • Aquarius will be re-working the Sheba’s Ridge Bankable Feasibility Study in the next few months, but this project will not be a

priority for the company for the immediate future. • Blue Ridge’s unit cash costs are expected to be similar to Everest’s, somewhere between Kroondal and Marikana.

Outlook • Anglo Platinum’s admission of its inability to mine Bleskop profitably we believe represents a golden opportunity for Aquarius to

strike a deal to extend Kroondal’s life-of-mine. • Milling is set to recommence at Everest by September 2010. • The Ridge Mining acquisition has helped improve Aquarius’ growth profile. Nevertheless, Blue Ridge is relatively minor in the

grand scheme of things (Aquarius’ share being half of Blue Ridge’s projected annual steady-state production of 125 koz 4E). Sheba’s Ridge promises to be much bigger, but we are not convinced that this project can be developed in the absence independent PGM smelting and refining capacities - something Aquarius is unlikely to attempt to develop alone, given its low-risk inclination.

• In short, we believe that Aquarius will continue to seek out potential projects with which to add growth.


7

Eastern Platinum

Crocodile River Mine RBC visited Eastern Platinum’s (TSX & AIM: ELR; “Eastplats”) Crocodile River Mine (“CRM”) on Tuesday, January 26th.

CRM Targeting 175 ktpm

Eastplats is aiming to maintain steady-state primary mine production at 175 ktpm (2010) at CRM, from H2-2012:

• Zandfontein Section – Zandfontein is already producing at 100 ktpm (120 koz 6E). The previous plan was to ramp up to 120 ktpm. However, such a rate could not be maintained consistently, which would have adverse plant processing effects; Eastplats has calculated that it would be better to maintain a consistent 100 ktpm production rate.

Exhibit 2: Conventional Breast Mining At CRM. We travelled down Zandfontein’s 450m-deep vertical shaft to visit a panel on 2-level 7-West. Working with stope widths of <1.5m, at dips of 17°-18° is not easy, and we believe that the workforce is doing commendably well.



8

• Maroelabult – Production is expected to be 35-40 ktpm (40 koz) for 2-3 years, before falling away (Maroelabult has around 10 years’ life-of-mine left at the moment).

• Crocette – Eastplats had just started sinking two declines at Crocette when the global downturn hit. Work was halted and the declines put on care-and-maintenance. In January 2010, Eastplats decided to restart development, and the declines are currently being dewatered. Development is scheduled to start in June 2010, with first development ore expected in November 2010. A 40 ktpm steady-state rate (50 koz) should be achieved in H2-2012. The capex for Crocette’s development is expected to be around R50 million and R150 million in 2010 and 2011, respectively.

• Kareespruit – Kareespruit could potentially double CRM’s production to 400 koz 6E annually. However, it would require sinking of deep shafts and new plant. Consequently, Eastplats views this as a long-term project, ranking lowest in priority of the various projects in CRM’s pipeline. There is potential synergy with Zandfontein, which itself will require a deep vertical shaft by 2018 (around 1.2 km deep) - a shared shaft system should save significant capex. Importantly, the shafts would be shallower than many comparable Majors’ shafts (often 1.6-2 km deep).

Exhibit 3: Crocette Requires Dewatering; Should Be Completed In H1-2010.


In the short term, CRM production should be supplemented by dump retreatment - potentially 10-15 kg (4-6 koz) per year. CRM has approximately 3 Mt of tailings, which at 80-100 ktpm should last for 3-4 years. The dump operations are making use of what had originally been the CRM plant’s tertiary circuit.


9

Exhibit 4: Dump Retreatment at CRM


Power No Problem At CRM CRM-sans-Crocette uses 18 MVA of its current 23 MVA allocation, and various power-saving programmes are being implemented. Meanwhile, Crocette is expected to use 2-3 MVA at full production, versus Marolelabult’s 4 MVA. Crocette’s lower power requirement is due to Crocette’s envisaged use of hydropower mining systems, à la Northam – Crocette will effectively be serving as a hydropower testbed for future vertical shafts at CRM.

In the short-to-medium term, therefore, power is therefore not expected to be a problem for CRM. The mine should cope even if South African power utility Eskom imposes a 10% cut in future.

CRM’s Recoveries Could Improve Still Further The previous maximum recovery achieved at CRM was 80%. We learnt during the visit that CRM had touched 81% (although it remains to be seen if this was a welcome blip or something that can be sustained). That said, there is clearly potentially scope to increase recoveries further (e.g., to 82% or beyond) for relatively little additional capex – we noted that Eastplats’ plant has currently-unused flotation cells that could be brought into the flow sheet.

Exhibit 5: PGM Flotation At CRM. The Plant Has Currently-Unused Flotation Cells.


Aiming For Q2-09 Costs CRM’s cash cost is currently around R500/t at moment (was R450/t before mid-09 strike). Additional volumes from Crocette should help reduce costs. Eastplats is hoping to reduce costs 10%, or with luck, 20% (although this would be rather optimistic).


10

Spitzkop & Mareesburg Studies Being Dusted Off Eastplats is currently updating the Spitzkop and Mareesburg feasibility studies; the updated studies could be completed over the next few weeks. Decisions will then be taken on development timelines and funding; Eastplats anticipated announcing these by June 2010.

• Fraction of New Cost - According to Eastplats, a brand-new 200 koz operation would cost around US$1 billion to build today. In contrast, Eastplats believes it could bring Spitzkop into production for US$200-US$250 million, partly because some work had already completed prior to the global downturn. Certain long-lead items such as crushers and mills are already on-site.

• UG2-Only Scenario Would Save On Capex - Eastplats currently favours mining only UG2 (80-90 ktpm; 100 koz 6E) at Spitzkop. The Merensky reef here is wide and there is unpredictability of where the grade peaks occur (so it would not be possible to selectively-mine a portion of the reef). Mining only the UG2 could reduce capex requirements by around 30%.

• Two Possibilities for “Spare” Capacity – If Merensky reef is not mined at Spitzkop, we believe the “spare” plant capacity could be used to treat Mareesburg UG2 ore (as opposed to having it toll-treated). Alternatively, or in addition, the Merensky decline could be re-routed to access UG2 at the De Goedeverwachting (“DGV”) portion of Kennedy’s Vale (another 18 months or so to reef).

• Better Than CRM? - The UG2 at Mareesburg and Spitzkop grades around 1-1.5 g/t higher than at CRM. Recoveries are also expected be better, circa 82%-85%.

• Water No Longer A Constraint? - The De Hoop dam appears set for completion next year, (unusually) a year ahead of schedule.

Production could start at Spitzkop around 18-24 months from a “Go” decision; i.e., possibly as early as the end of 2011. Consequently, Eastern Limb production rates could potentially reach 250 koz in 2012 (Spitzkop and Mareesburg) and >300 koz rate in 2013 (>500 koz including CRM!).

Exhibit 6: Merensky and UG2 Portals At Spitzkop. When the project was put on care-and-maintenance, the Merensky decline had already hit reef; the UG2 decline was three months away from reef.

Source: Eastern Platinum

Chrome Production The spiral sections in CRM’s plant are allowing concentrate chrome levels to be maintained at 1.8%-2%, resulting in relatively-low chrome penalties.

The chrome spirals do not merely minimize chrome penalties – they also generate handsome revenues and profits. CRM is currently producing UG2 chromite concentrate at a rate of around 320 ktpa at the moment; revenues are around R450/t, versus production costs of around R25-30/t(!).

We expect that Spitzkop-Kennedy’s Vale and Mareesburg should boast similar levels of chromite production and profitability.


11

Exhibit 7: Chrome Spirals at CRM. Unusually, CRM has spirals incorporated into both primary and secondary circuits.


Status Of Key Mining Rights Eastplats has New Order mining rights for CRM and Spitzkop. A mining right for Mareesburg is expected shortly as the Social and Environmental plans have been approved.

The Next Aquarius? Eastplats has survived the recent global economic crisis in good condition. No dilutive equity placings were done or expensive debt taken on, CRM is constantly breaking records, costs are relatively low, chrome is proving a very nice sweetener and there is plenty of organic growth potential (possibly 500 koz by 2014).

Eastplats has, however, depleted its previous cash pile and currently has around C$30 million in working capital. Spitzkop and Mareesburg could be funded from internal cash flows, according to our model. However, given the rise in its share price, we believe Eastplats could well be tempted to come to the market to speed up these projects.

In short, things are generally looking rosy for Eastplats. In our view, Eastplats could well be the next Aquarius – provided, of course, that the relatively low-capex and low-cost nature of its large (100 Moz) resource base doesn’t result in the company being taken out first.


12

Jubilee Platinum

Mintek 3.2 MV PGM Smelter RBC visited Jubilee Platinum’s (AIM: JLP; “Jubilee”) 3.2 MW smelter at Mintek on Tuesday, February 2nd. Appropriately enough, the weather was roasting hot...

What Is ConRoast? • The ConRoast process comprises the sulphur-free reductive smelting of PGMs into an iron-alloy collector in a DC Arc Furnace. • Conventional PGM smelters use AC technology (not DC) and smelt into a sulphide matte (as opposed to an iron alloy). • ConRoast can be thought of as comprising three discrete stages: roasting, smelting and refining. In our view, the first two can be

regarded as proven. Refining options are currently still being studied and finalised, but this final step is not absolutely crucial (see below).

Exhibit 8: Top Row - Furnace Tapping; Bottom Row - Alloy Ingots, Crushed and Granulated.


ConRoast Addresses A Variety Of Current Smelting Issues ConRoast offers a number of benefits to Major and Junior PGM companies (although we point out that many of the benefits to Majors would also flow through to the Juniors, as will be seen):

• Majors - ConRoast effectively overcomes a number of problems plaguing existing conventional PGM smelting operations. Key ones include: 1) Constraints on Furnace Feed Chrome Levels - The proportion of PGM production from high-chrome UG2 sources is ever-

increasing as viable Merensky reserves are depleted. Conventional AC furnaces are limited in their chrome tolerances - chromite spinel buildup reduces furnaces’ volumes and traps PGMs in slag, reducing furnace PGM recoveries. DC technology suffers no such constraint, as conditions in DC furnaces result in chrome dissolving readily in the slag.

2) UG2 is Sulphide-Deficient - As mentioned above, UG2 production is only ever going to increase. UG2 is generally relatively-deficient in sulphides. This is a problem for traditional furnaces, which utilise sulphides as a collector. In contrast, DC technology uses an iron alloy collector.

3) SO2 Emissions - Limiting SO2 emissions in the face of ever-tougher legislation is much tougher with traditional AC technology. ConRoast utilises a roaster ahead of smelting and refining, allowing SO2 to be captured more efficiently, facilitating much-lower SO2 emissions.


13

4) Material Containment - In AC furnaces, highly-corrosive superheated matte can burn through furnace walls, leading to smelter runouts - in DC furnaces, there is no such danger as there is no need to subject material to temperatures that result in superheating.

In overcoming these problems, ConRoast allows “difficult” high-chrome materials to be processed, e.g. UG2 concentrates and smelter reverts. Additionally, PGM recoveries into UG2 concentrate can be increased as there would no longer be a need to try to minimise concentrate chrome content (efforts at which tend to reduce PGM recoveries).

• Juniors - Most Junior projects are based on UG2 reef and/or low-grade resources which yield low-grade PGM concentrates (the processing of which would displace higher-grade concentrates at Majors’ furnaces, an opportunity cost). Given Majors’ smelting issues, finding a home for concentrates from such sources could be a problem in future. Alternatively, or in addition, Majors could impose more onerous tolling terms, and/or more punitive penalties where concentrate specifications are breached.

1) Installation of DC technology at Majors’ operations would circumvent the chrome issue (see above). As mentioned above, PGM recoveries into UG2 concentrate can be increased as there would no longer be a need to try to minimise concentrate chrome content. However, where chrome spirals have been installed, this could mean little or no UG2 chromite concentrate produced. Clearly, companies would have to balance higher PGM recoveries against reduced chromite sales revenues; hypothetically, if chromite prices rose enough, it might be better to sell more chromite and less PGMs.

2) The establishment of “alternative” junior toll smelters (e.g., Jubilee) promises an alternative home for low-grade concentrates. That said, capacity could be limited - this would be dependent on how many furnaces Jubilee builds, or licences others to build. Limiting DC capacity (own or out-licensed) could actually be advantageous to Jubilee in potentially allowing Jubilee to extract higher tolling or price participation terms (e.g., Jubilee could well pay Junior companies <80% of the contained metal value in their products if the Junior companies are desperate). Consequently, ConRoast is by no means a complete panacea for Junior PGM companies.

Upstream And Downstream Choices • Roasting Not Always Necessary - Roasting is not required if the furnace feed has little or no sulphide content (e.g., PCM

concentrates derived from retreated chrome tails). Consequently, if it is known that a furnace’s feed will comprise such materials, build capex savings can be achieved by deferring or not installing the roaster.

• Downstream Choices Aplenty - The other key choices are to do with route and degree of refinement (if any) of the smelter product. A DC smelter can opt to sell untreated PGM-containing iron alloy, or (semi-)refined products. Generally, the more refined the final products, the better their payability; also, the more potential customers for a particular product type, the better terms are likely to be.

• Two Main Downstream Routes - There are two main refining routes: pyrometallurgical or hydrometallurgical. The former is simpler, using existing converter technology; the resulting semi-refined product can be further processed in existing base metal refineries. The latter is more complicated, requires more development work (there are two major sub-routes) and more expensive but could potentially yield more and cleaner products, attract a more diverse range of clients and boast better payability.

We stress that the fact that development work is still on-going on the hydrometallurgical route is not the hurdle or problem that some might suggest. As pointed out above, the absolute worst case scenario is that smelter matte is sold on to refiners at a steeper discount. More likely, converting will be used in the short-term, and converter matte is in fact an attractive product for existing PGM refineries.

In short, whether roasting is installed is dependent on what materials a DC furnace is intended to treat. Meanwhile, refining choices will depend on progress in developing hydrometallurgical refining options and on customer demand for (semi-)refined products.


14

Alternative “Third” Refining Route Could See Much Higher Payability Jubilee’s vision for the Middelburg PGM toll-smelter has changed significantly over the last half year or so. In mid-2009, our impression was that Jubilee would go for a “cheap and simple” furnace-&-converter setup, with the whole facility requiring a maximum of 10 MW. • Management is now leaning towards a 16 MW facility (10 MW furnace, operated at 8 MW) including roaster, converter AND

refining facilities (henceforth, “refining facilities” refers to non-converter routes). Part of the 6 MW capacity for auxiliaries is actually “reserved” for a refinery upgrade - Jubilee anticipates expansion to process alloy from the Jubilee-Sylvania joint venture furnaces on the Northern Limb (see below).

• It emerges that a “third alternative” refining route (i.e., not converting or leaching) has apparently shown significant promise in pilot plant testwork to date. Not only will the PGMs have much higher payability (we believe possibly up to 97%), but this “third” process produces pure nickel and iron, which will have 100% payability. Testwork should be completed in March or April 2010.

• A roaster will be included as sulphide-bearing materials are likely to be treated (previously, the thinking was to focus initially on non-sulphide-containing materials such as chrome-derived concentrates). The converter is being kept in the process sheet as backup (e.g. during refinery downtime) and/or for treating certain PGM-bearing materials.

Changes And Other Issues Mean Delay, Higher Costs • Middelburg Capex Now Much Higher - The inclusion of a roaster and refinery, plus various additional costs (such as site

refurbishment, security, etc) mean that Jubilee’s capex estimate has ballooned from R200-R250 million to R420 million excluding the refinery, or R610 million including the refinery.

• Refining Opex Should Be Offset By Higher Payability - Opex is estimated at R1300/t feed for roasting-plus-smelting; refining could cost around the same again, but it is very early days to be trying to put firm figures to refining opex since the flowsheet hasn’t been finalised (note: this is also the case for refining capex).

• Eskom Delays - There have also been delays stemming from negotiations with Eskom (which is to provide 6 MW; the remainder is to be self-generated from gas supplied by Sasol) and with the owner of the Middelburg brownfield site.

The sum of the above is that a final build decision is now expected only towards the end of H1-CY10, with commissioning now targeted for Q4-CY11.

More Funds Needed For Enlarged Scope • Insufficient Funds… - The much higher capex requirement means that the monies originally raised in 2009 for the “simple”

smelter will now be insufficient - Jubilee currently has around R240 million in the bank. • …but Partner Could Contribute - We understand that Jubilee will either get its “Middelburg partner” (unspecified major, which

will be providing most or all of the feed) to contribute in return for some sort of interest or stake, or come to the market for more funds.

Opex Not Key Concern In Jubilee’s Case • Smelting unit costs will be dependent on furnace size, with smaller furnaces generally being less efficient and more costly to run in

terms of electricity utilisation. • DC furnaces being planned are generally smaller than existing AC ones. They are therefore likely to be slightly less efficient in

electricity usage, but will generally save on other operating costs (e.g., electrodes, fluxing). • For practical purposes, we recommend the assumption of similar unit costs for AC and DC furnaces.

More to the point, we believe an over-emphasis on unit smelting costs to be inappropriate. • What a small DC furnace might lose in higher unit costs should be at least (and probably more than) made up for in better

recoveries at concentrator and smelter levels (and potentially, better sulphuric acid sales). • Larger numbers of smaller furnaces offer better operational flexibility. This is a not unimportant consideration given the electricity

situation in South Africa. More, smaller furnaces could actually run more efficiently and produce better cost performance in times of electricity cuts and outages.

• In any case, we expect that Jubilee would impose on at least a portion of materials treated (certainly low-grade feeds) a fee structure that recoups its processing costs. (Such a fee structure would cap both downside and upside for a DC smelter, so we expect that Jubilee would negotiate a payability regime for higher-grade materials in order to participate in any price upside).


15

Power Availability More Critical • Power availability currently appears the key constraint on Jubilee expanding its toll smelting business, but this could actually work

in Jubilee’s favour (in allowing the company to charge more for toll processing or pay less for feed, and/or to be more selective in the feed grade it accepts).

• As alluded to above, South Africa’s electricity crisis could well come back. In times of electricity cuts and outages, larger numbers of smaller furnaces (e.g., three 10 MW furnaces rather than one 30 MW furnace) could actually run more efficiently and produce better cost performance.

• Brownfield sites may have existing power allocations that can be used. A variation on this theme would be to convert or replace existing AC PGM or chrome furnaces with DC furnaces.

Jubilee’s ConRoast Rights Appear Robust Jubilee holds exclusive rights to the Mintek ConRoast patent until 2020. A number of companies are trying, or have tried, to circumvent this patent - but none with any real success to date:

• The ConRoast patent still belongs to Mintek - effectively, the South African government. Thus, any company attempting to circumvent the ConRoast risks incurring the wrath of the South African government.

• Barrick had tried to register a similar patent, but backed down when made known of Mintek’s patent. • Sylvania Resources abandoned attempts to develop alternative technologies and recently announced joint ventures with Jubilee

instead. • No viable legal alternatives to ConRoast have emerged in the last decade; existing workarounds have tended to be uneconomic,

plagued with technical challenges or indeed fail to address the technical issues that ConRoast solves.

In short, all the evidence to date suggests that the ConRoast patent is robust indeed.

One Significant Deal Already, More On The Cards • As noted above, Sylvania Resources has abandoned attempts to develop alternative technologies and is currently concluding a

30:70 smelting joint venture with Jubilee instead (plus a 50:50 tailings retreatment joint venture). The smelter joint venture will see three 10 MW furnaces built to process concentrate from Sylvania’s Northern Limb properties.

• We understand that the Majors and Northam have all contributed material for testing in the past. Of these, we see Northam as the most likely to do a deal in the short-term - Northam requires additional capacity to process Booysendal’s production and offtake from Platmin; the preponderance of UG2 production from Northam mine and Booysendal makes DC a logical choice.

• Of the Majors, we see Lonmin as most likely to do a deal. Indeed, Lonmin has publicly stated that it is looking at smelter backup capacity options, to reduce the company’s dependence on the No.1 furnace. Lonmin has a particularly high proportion of UG2 production and is short on sulphide sources for the time being.

• Impala has stated plans to increase production at Zimplats to 1 Moz in the long-term. We believe that DC technology would be considered in expanding the Zimplats smelter. In addition, Impala may well be tempted to add a DC furnace to its existing IRS operations, given that Junior toll feeds will be increasingly from UG2 sources while Impala’s Rustenburg mine has problems with Merensky availability.

• Depending on the state of the ferrochrome markets, ferrochrome producers might well consider converting ferrochrome furnaces to PGM DC ones or building new PGM DC ones, or simply “leasing” any unused power allocations to Jubilee (so as not to “lose” their power allocations).

According to Jubilee, another Sylvania-type deal may be announced this quarter or next. We caution that one can never be sure how long negotiations will last, but this is good news nonetheless.

Conclusion - Exciting Times, As More Come To The Pot ConRoast has come far enough that there can be no doubt of the technical viability of the roasting and smelting stages.

• That hydrometallurgical refining is not yet proven to the same extent is not a major headache; converting offers a viable alternative, albeit with slightly-less value add.

• Jubilee’s ConRoast rights appear secure, as evidenced by Sylvania’s decision to halt work on alternative technologies and secure a deal with Jubilee instead.

• The delay in getting the Middelburg smelter built and the potential for Jubilee having to come back to the market will be disappointing to some.

• There is considerable potential for deals with other companies, and Jubilee believe one could potentially be announced in Q1 or Q2. • Power availability is the key constraint on Jubilee expanding its toll smelting business, but this could potentially work in Jubilee’s

favour (e.g., Jubilee can charge more for toll processing or pay less for feed).


16

Platinum Australia

Rooderand PGM Project RBC visited Platinum Australia’s (ASX: PLA; AIM: PLAA) Rooderand project on Saturday, January 30th.

PLA Earning Into 70% In May 2009, PLA announced a deal with Bakgatla-related BEE company Atla Mining Resources (“Atla”) on Portion 2 of farm Rooderand 46JQ (“Rooderand”). Rooderand is located just to the north of the Pilanesberg Alkaline Complex, immediately adjacent to Platmin’s Tuschenkomst open-pit. • PLA earned an initial 30% interest by paying R13.5 million (A$2 million) after a New Order prospecting right was issued to Atla. • Another 35% will be earned by completing a Definitive Feasibility Study (“DFS”; funded by PLA). • A further 5% can be earned by PLA’s arranging financing project for project construction.

Exhibit 9: Note Heavy Faulting At Tuschenkomst (Yellow Triangle) and Platmin’s Portion of Rooderand (Bottom); Pink Oblong Roughly Denotes PLA’s Portion 2

Source: Platinum Australia, Platmin


17

Ambitious Work Programme • PLA commenced drilling in late November 2009, just after the New Order prospecting right was received. The plan is to have 100

holes drilled by Q1-FY11; to this end, a third drill rig is to be added in Q3-FY10. • By the time of RBC’s visit, 21 holes had been drilled. • A Pre-Feasibility Study (“PFS”) is to commence in Q3-FY10. Completion of the Definitive Feasibility Study (“DFS”) is targeted

for Q1-CY11.

Potentially Bigger Than Smokey Hills PLA believes that Rooderand could produce 150 koz 4E annually for 20 years, versus Smokey Hills’ 95 koz for 7 years.

• Both Merensky and UG2 reefs outcrop on the property and descend to around 500m. • Previous owner Anglo Platinum delineated Inferred UG2 resources of 11.8 Mt of @ 5.39 g/t 4E (2 Moz) and Merensky resources

of 3.8 Mt @ 7.99 g/t (1 Moz). • Rooderand is stratigraphically very-similar to Platmin’s Pilanesberg; early drilling results have already indicated the presence of

Merensky, Pseudo and UG2 Reefs. • Where fully-developed, this package can be 20m-30m thick; it should be noted that much (if not all) of Tuschenkomst comprises

the narrower “Contact” and “Pothole” facies. We would therefore be unsurprised if this also proved the case at Rooderand.

High-Grade Strategy • Should the go-ahead be given for Rooderand, PLA currently favours processing only high-grade material. • Low-grade material (likely a significant portion of the Pseudo reefs) would be stockpiled for future processing, if PGM prices are

sufficiently favourable.

Consolidation Makes Sense Nkwe and Platmin own portions of Rooderand to the west of Portion 2, while the Bakgatla own the portion to the east. Additionally, we believe that it would make sense in the longer-term to extend Platmin’s Tuschenkomst pit into the northern portion of PLA’s Rooderand.

• In our view, this raises the intriguing possibility of consolidative transactions between the parties, which would potentially create a larger, longer-life operation with lower unit costs.

• More ambitiously, it would seem logical to combine the Rooderand portions with Platmin’s Pilanesberg, what used to be Barrick’s Sedibelo (immediately east of Rooderand and Tuschenkomst) and Magazynskraal (immediately east of Sedibelo). A larger combination of properties could well justify a standalone smelter complex.

• Another possibility that suggests itself would be some sort of ground swap. For example, PLA could exchange Portion 2 for Platmin's Grootboom. Grootboom is very close to Smokey Hills, and could potentially be used to increase Smokey Hill’s life of mine and/or annual production.

RBC Expectations • We expect an initial open-pit operation, followed by underground mining should the geology prove amenable (i.e., ground not

overly broken-up by faulting). • Assuming a mining right is received by early FY12, we believe that production could start by mid-FY13.


18

Platmin

Pilanesberg Platinum Mine

A View From The Fence Platmin (TSX & AIM: PPN) is not entertaining visitors at the moment, focusing on righting issues plaguing the ramp-up at Pilanesberg. • There has been much market chatter about the exact nature of the underlying problem(s), but most appear to concern the DMS. • One line of speculation has been that new CEO Tom Dale has commanded the halting of bulk (low-grade) siliceous mining,

focusing on high-grade material and cutting out the DMS. • On Saturday Jan 30th, we passed the Pilanesberg plant. The Merensky, UG2 and DMS plants all appeared to be operating normally,

so the above contention would appear to be incorrect, in our view.

Exhibit 10: DMS Clearly Operating



19

Sylvania Resources

Samancor Dump Operations RBC visited the Millsell and Mooinooi plants of Sylvania Resources’ (AIM: SLV; “Sylvania”) Samancor Dump Operations (“SDO”) on Monday, February 1st.

Background Sylvania’s 74%-owned SDO recovers PGMs from chrome tailings and current arisings from Samancor Chrome’s mines on the Bushveld Complex’s Eastern and Western Limbs.

• Essentially, Sylvania produces (at least a minimum amount of) chromite concentrates for Samancor at cost, and takes the PGMs as “payment”.

• SDO started with two plants, Millsell and Steelpoort, which came into production in FY08. These were followed by two further plants in FY09, Lannex and Mooinooi. Two more are in the works: Doornbosch and Tweefontein.

Mooinooi – Ramping Up Mooinooi, located at Samancor’s Western Chrome Mines some 30 km east of Rustenburg, was commissioned in December 2009. It is currently in ramp-up, with design production to be reached in late Q3-FY10.

• Feed sources potentially include the Mooinooi, Elandsdrift and Buffelsfontein tailings dams, and current arisings from Samancor’s Mooinooi plant. Mooinooi will also receive ROM feed from the Buffelsfontein Mine and Mooinooi shaft.

• The 37 ktpm plant receives feed at an average feed grade of 1.2 g/t 4E, and planned production is currently around 6.8 koz 4E @ <US$350/oz 4E.

• Sylvania is planning to install bead mills in the near future. It appears column cells will be unnecessary, due to the coarser nature of the chrome here – Sylvania expects to be able to produce concentrates grading >160 g/t with <3% chrome.

The Mooinooi dump rights were actually purchased from Lonmin in a recently-concluded transaction. Consequently, Lonmin has first right of refusal on PGM concentrates from Mooinooi. However, given Lonmin’s problems with its No.1 furnace and Merensky shortage, we believe Lonmin is unlikely to take up its rights. Indeed, we understand that Mooinooi’s concentrates are currently being toll-treated by Anglo Platinum.

Exhibit 11: Mooinooi Plant.


Millsell – Steady As She Goes Millsell, located at Samancor’s Millsell plant some 10 km southeast of Rustenburg, was Sylvania’s first SDO plant.

• Feed sources include the Millsell tailings dam and current arisings from Samancor’s Millsell plant. • The 37 ktpm plant receives feed at an average feed grade of 1.8 g/t 4E, and planned production is currently around 9.5 koz 4E @

<US$350/oz 4E. • Millsell was apparently the only one of Samancor’s mines that did not halt production during the recent global downturn.


20

Exhibit 12: Millsell Plant.


Joint Ventures With Jubilee Announced In early February 2010, Jubilee Platinum (“Jubilee”) and Sylvania agreed to incorporate two joint venture companies within three months: • The Smelting Company (“SmeltCo”) – SmeltCo’s ownership will be 70:30 Jubilee:Sylvania, and managed by Jubilee (which will

receive a profitability-related management fee). • The Tailings Processing Company (“TailsCo”) – TailsCo’s ownership will be 50:50. The company will ostensibly be jointly-

managed, but we understand that Sylvania will receive a management fee.

SmeltCo SmeltCo is to use Jubilee’s proprietary ConRoast technology to smelt and refine concentrates from Sylvania’s Northern Limb properties. SmeltCo will also receive any uncommitted concentrate from Sylvania’s existing tailings operations, and TailsCo concentrates.

• Subject to a “Go” decision being taken on Sylvania’s Northern Limb properties (expected later in H1-CY10), SmeltCo is expected to commission the first of two or three 10 MW ConRoast furnaces (with associated refining and other ancillary facilities) within three years.

• Sylvania will receive for its offtake a revenue price not less than that it currently receives from the Majors.

TailsCo TailsCo is intended to take advantage of any new tailings retreatment opportunities; i.e., Sylvania’s and Jubilee’s existing assets are excluded.

• The logic behind TailsCo is that it provides the PGM industry with a “total solution” for the processing of high-chrome and/or low-grade PGM concentrates, which should allow TailsCo to compete for tailings resources more effectively.


21

Central Rand Gold

CMR Trial Mine Positive Initial Trial Mining Results Central Rand Gold (LSE: CRND; “CRG”) Site Visit, Sunday January 31st.

RBC visited Central Rand’s CMR trial mine on Sunday, January 31st. Coincidentally, this was the first truly sunny, rain-free day in whole of the last week of January 2010.

Exhibit 13: Drilling A Strike Gully To Mine Main Reef (Top Row); Note Main Reef Leader Voids (Bottom Row).

Source: Central Rand Gold

Trial Stoping Just Started But Mining Already No Longer The Key Concern • CRG has started “trial mining” in “Trial Stoping Block #1 (Block 1), a longhole stope intended to test the new geotechnical

support system (“Cemented Aggregate Fill”, or “CAF”); longhole stoping in Trial Stoping Block #2 (“Block 2”) intended to test the mining itself is to be conducted in February 2010 (10 days from today), although the strike gullies have been prepared.

• One panel in Block 1 has already been blasted and filled with CAF. Although still early days, Snowden (and CRG) was already happy that this demonstrated that longhole stoping and the CAF system generally works.

• Furthermore, although the panel was taken in a single blast, Snowden is of the opinion that two-blast resue mining will not be any more problematic.

• Strike gully development also demonstrated CRG’s ability to develop through areas where the Main Reef Leader (“MRL”) hanging wall had come down.

• We caution that work still needs to be done on the geotechnical side (mainly to demonstrate that CAF pillars will survive blasting of panels in between) and on stoping parameters such as dilution and ore modifying factors.

• Trial mining should be completed by the end of Q1-CY10.


22

Implications Of The Initial Results In short, CRG has demonstrated that the mining and support methods proposed are practical; the questions that remain mainly surround completeness of extraction, degree of dilution and exact cost of mining. That said, note that:

• CAF is cheaper and much stronger than the previous support system envisaged, • Longhole mining offers the potential for selective extraction of higher grade bands as thin as 50cm (or alternatively, a combination

of bulk mining with optical sorting), and the potential to collect MRL sweepings and vampings at the same time as middling. Recent drilling has identified the presence of mineralisation in the middling and trial stoping has revealed bands of mineralisation in the footwall.

• The combination of CAF and longhole stoping suggests the possibility that MRL pillars (which carry very high grade) can be extracted also.

• Mechanised development and mining suggest low costs; the high degree of mechanisation also means very few people in stopes and much lower chances of injuries and fatalities (important given the degree of disruption Section 54 stoppages can cause).

Geologists And Access - Keys To Mining Success The keys to (degree of) success with CRG’s longhole methodology are:

• Access (decline driving) - which will determine how many attack points can be exploited at any given time. The “new” contractor, ACM from Australia, is achieving rates of 60m/week, versus the 60m/month achieved by CRG’s South African ex-contractor Grinnaker,

• Geologists - Chip sampling will be undertaken every 2m to construct vertical grade profiles; grade will direct mining (resue or whole-blast).

Optical Sorting And CIL Plant - Keys To Processing Success • Wherever ore is bulk mined (i.e., single blast), it will generally be subjected to optical sorting, which will create two streams: a

high-grade one and a low-grade one. (The key exception being that fines will go direct to the CIL plant.) Trials on a low-grade 30t bulk sample at Mintek were successful.

• High-grade ores are to be subjected to “direct” CIL processing, which has higher recoveries (c95%) but also higher cost/t processed.

• Lower-grade material will be floated (lower recovery - c85%, but lower cost/t processed) and the concentrate sent to the CIL plant. • CIL plant processing is significantly higher-cost than concentrator processing. Putting lower-grade material through the

concentrator first reduces the tonnage that the CIL plant has to treat, resulting in overall cost savings. • In addition, overall gold recoveries should be improved (from 80% to 86%) as much less gold is lost in the concentrator tails. CRG

is proposing to triple CIL capacity to 30 ktpm (US$11.5m), which would increase overall recoveries from 86% to 95%, which could potentially allow the concentrators to be largely or even entirely bypassed.

• The difference in recoveries suggest that where there is a significant grade difference, reef or mineralised bands should perhaps be mined selectively; conversely, where there is not, a single-blast approach could be applied.

Main Reef Upside Potential Aplenty • According to CRG, results to date appear to confirm the robustness of CRG’s data - aside from at the very start, when a couple of

unexpected voids were encountered. • CMR East is lower-grade than CMR West, so it was encouraging to see that there is a very-rich payshoot (including a significant

amount of mineralisation near or at surface) within the mining right area, at Crown. This is currently only an Inferred resource; a drilling programme is planned for later in CY10.

• The catch is that the Crown payshoot area is more densely populated, which could make drilling tricky. • There are also apparently-rich areas in the Robinson/Village tenement, but again, there are more industrial/residential structures.

Furthermore, Robinson/Village is currently held under a prospecting right, rather than a mining right. • CMR is in fact probably the lowest-grade of CRG’s properties, but the company started here due to the lack of surface structures.


23

Joint Ventures For Other Reefs? • Other companies have apparently approached CRG about the possibility of prospecting other reefs, notably the Birds and

Kimberleys, for gold and uranium. • If CRG agrees, CRG would probably just take royalties on production (circumventing the need to raise capital and the effort of

building mines).

Outlook • Given the above, CRG is targeting a production rate of 55 koz (48 ktpm) by CY11. • According to Snowden and CRG, things are still on track for a revised CPR to be issued in Q2-CY10 (March-April). It should be

noted that the revised CPR will still NOT give CRG any grade in the middling, footwall, sweepings & vampings or other reefs. Post-CPR, this could change once such mineralisation is included in CRG’s resource & reserve statements.

• A fund raising will then follow, currently envisaged at CRG to be in the vicinity of US$25-35 million. This includes US$10 million for a pump station but does NOT include an optical sorter or CIL upgrade.

• Given trial stoping results to date, should trial processing prove successful, we would be unsurprised to see the fund-raising figure rise to around US$45-50 million.

• Given the initial trial mining results, CRG’s future is now looking decidedly more upbeat. That said, we caution that it is still early days - we look forward to further positive trial results, particularly with regards blasting between CAF pillars and optical sorting.


24

DRDGOLD

ErgoGold RBC visited DRDGOLD’s (JSE: DRD) ErgoGold (“Ergo”) operations on Thursday, January 28th.

Ergo was founded to retreat up to 1.7 billion tonnes of surface tailings near Benoni, Springs and Brakpan (all located to the east of Johannesburg), primarily to recover gold.

Over 2008-2009, DRD acquired Mintails’ 50% in Ergo’s Phases 1 and 2 for R277 million, giving DRD full ownership of Ergo’s Brakpan plant, doubling available capacity and providing DRD full exposure to Phase 2’s uranium and sulphuric acid potential.

Exhibit 14: Ergo’s Brakpan – A Giant-Scale Gold Plant.



25

Ramping Up Phase 1 Steady-state production from the gold-only Phase 1 is to be 1.2 Mtpm at an average head-grade of 0.32 g/t and at a cost of R30/tonne. • In Q2-FY10, gold production rose >80% to 9.45 koz, reflecting a 34% throughput build-up to 3.124 Mt (or 1.04 Mtpm) and a 29%

improvement in average yield to 0.09 g/t. • This resulted in cash costs falling 33% to R223,224/kg.

Exhibit 15: Mining The Elsburg Tailings Complex.


Benoni Dump Continues To Be Troublesome, But Not For Too Much Longer Volumes from the L29 (Benoni) dump have been satisfactory, but recoveries (as low as 0.02 g/t at the start!) have not. • The Benoni dump has previously been re-mined and as a result contains an abnormally high amount of fine preg-robbing carbon. • In August 2009, production commenced from the Elsburg Tailings Complex (“ETC”) via the second 0.6 Mtpm feeder line. This

has been helping increase head-grade, recoveries and therefore gold production - DRD is essentially diluting Benoni material with ETC material; a pilot plant was constructed to help determine the optimal blend.

• In short, DRD has not been able to truly resolve the L29 problem, and head-grade and recoveries are likely to hit specifications once Benoni’s retreatment is completed (around 11-12 months’ production left).

That said, DRD is continuing to investigate technologies with which to better re-retreat dumps such as Benoni. In addition to improving current recoveries and profitability, such technologies could potentially unlock vast resources in the form of dumps that DRD has retreated.

Exhibit 16: Ergo Pilot Plant.


Royalties – Payable Or Not? DRD has applied for conversion of its mining rights from Old Order to New Order ones, but these conversions have yet to take effect. • According to DRD, the royalties payable in relation to the Mineral and Petroleum Resources Royalty Act are in fact only payable

on New Order mining rights. • In any case, dump mining does not currently require a mining right (although we believe this situation is likely to change in future),

and DRD’s position is that Ergo should therefore not have to pay any royalties (unless legislation is amended).


26

Water Cost To Rise DRD has adequate water for its requirements at the moment, but will have to start purchasing from external sources towards the end of 2010, which will add to costs (currently 3.5% of costs).

• That said, DRD should be able to source relatively-cheap water from de-watering operations at ERPM, which could potentially be available from FY12 – but this is dependent on the South African government agreeing plans proposed by the CBEC consortium.

• We understand that the government has been dragging its heels. However, given that water levels become critical in 18-24 months, the government does not have much room to prevaricate.

Ergo Excess Deposition Capacity Is Crown Opportunity Deposition space is not an issue for Ergo – there is 200 Mt of available capacity at the Brakpan tailings dam, and potentially another 300-400 Mt on an adjacent reclaimed area. There is also the Daggafontein facility, which has 150 Mt available.

• DRD is investigating potential synergies between Ergo and DRD’s Crown operations, particularly given Crown’s issues with depositional capacity.

• A R290 million pipeline linking the two is being mooted – DRD is intending to piggy-back on railway land, as this is more direct and therefore results in substantial capex savings over other routes. Total project capex would be R345 million (the additional capex being required for refurbishment of Brakpan’s second CIL circuit and for upgrading of tailings dams).

• This project was one of the motivating factors behind the sales of 60% of Blyvoor to Aurora for R296 million, and of the ERPM plant for R20 million.

Aurora Dump Opportunities Would Require Fund Raising • Aurora has paid R5 million towards the ERPM plant purchase and agreed to allow DRD to conduct a feasibility study on Aurora’s

Grootvlei and Marievale tailings dams near Ergo. • These dams contain 117.2 Mt (1.048 Moz) of Inferred resources; their treatment would require capex spend of R260 million. • DRD may choose to receive the outstanding R15 million in cash; alternatively, DRD could take an exclusive 10-year option to

retreat those tailings dams.

Should DRD exercise the Aurora tailings option, we believe that DRD debt or equity financing will be required, given the company’s December 2009 cash balance of R162 million.

Phase 2 Would Also Require Fund Raising Ergo’s Phase 2, which could also see uranium and sulphuric acid production, is at feasibility study stage.

• We believe that Phase 2’s capex would be in the order of several billion rand, so DRD would almost certainly have to raise equity and/or debt financing should this project be given the go-ahead.

Conclusion – Surface Production Is Preferable To Underground Ergo’s surface retreatment operations are:

• Low-technical risk, • Low in labour requirements (less chance of safety-related stoppages), • Low in power requirements (as compared to deep mines; important given the expected steep rises in electricity tariffs over the next

few years), and • At steady-state, should be relatively-low cost. Given the above, we see DRD’s rising proportion of surface production as a good thing.


27

Gold Fields

South Deep Gold Mine Gold Fields’ (JSE: GFI) South Deep is a developing, trackless mechanised mine situated some 45 km southwest of Johannesburg. It has two shaft systems (the Twin shaft and South shaft complexes) and a single plant. The mine has been subdivided into “above current infrastructure” (to 110 level; 2,888m) and “below current infrastructure” (to 135 level; 3,250m).

South Deep exploits the Ventersdorp Contact Reef (“VCR”) and the Upper Elsburg reef package (which subcrops against the VCR) at depths of 1,575m-3,500m. • The mine is unusual amongst South Africa’s deep-level gold mines in being a fully-mechanised operation. • The Main Shaft currently has a 175 ktpm capacity; completion of the Vent Shaft will add 195 ktpm, for a total of 370 ktpm (330

ktpm reef). • GFI is intending that South Deep will be producing at 330 ktpm, or 750-800 koz annually, by mid-FY15.

RBC visited the 95 level (around 2,600m below surface) in the 3-West corridor at South Deep mine on Friday, February 5th. The scale of mining and equipment was hugely impressive, particularly given the operating depths and despite that the visit (disappointingly) did not include any long-hole stopes.

Exhibit 17: South Deep’s Twin Shaft Complex. The Main Shaft boasts the world’s longest single-drop (2,995m); the winder is the world’s largest.


Vast Reserves South Deep has reserves of 29.5 Moz @ 6.1 g/t and resources of 63.8 Moz @ 7.2 g/t (excluding 14.6 Moz in the Uncle Harry’s ground). • Life-of-Mine (“LOM”) is estimated at 43 years; i.e., the current reserves are scheduled to deplete around 2052. • The Upper Elsburg package comprises 93% of the reserves versus the VCR’s 7%.

Mining Methods To Suit Orebody Characteristics The Upper Elsburg package is a wedge that thickens towards the east (up to 130m). In the west, the entire package is being mined (including internal waste partings). Towards the east, mining of selected reefs will become possible. Accordingly, a variety of mining methods will be used as appropriate: • Longhole Stoping – Cheapest mining method, applied to thickest portions of the orebody (>10 m cut, mainly to the east). Currently

responsible for 20% of tonnage; to reach 60% eventually. • Drift-and-Fill – Thinner portions in the west. • Modified Drift-and-Bench – Intermediate orebody thickness (5m-10m mining cut).

The Upper Elsburgs sit at around 2,500m-3,500m. Mining such wide cuts at such depths requires that in-situ rock stresses be reduced. This is achieved by “De-stress Mining”, which involves mining and backfilling a 2m slice through the Elsburg package. A de-stressed window of 50m-60 m above and below is created in which rock pressures are reduced from 80 MPa to 30-40 MPa (equivalent to stresses at 1200m).


28

Capex & Opex • GFI estimates an all-in production cost (“NCE”) at steady-state of R150,000/kg, or around US$625/oz. • In Q2-FY10, 71 koz was produced at an NCE of R394,000/kg (US$1,640/oz) or cash cost of R192,000/kg (US$800/oz). • Capital expenditure for FY10 is estimated at R1.7 billion. Over FY10 thru FY14, R8.4 billion will be spent.

Impressive Recent Safety Performance • In FY09, South Deep recorded no deaths at all. By December 9th, South Deep had achieved 560 fatal-free days (when sadly, a

miner was killed in a locomotive-related accident). • Given the potentially-significant economic impact of safety-related stoppages, it has to be hoped that South Deep is able to

maintain its low fatality rate (generally attributed to the mine’s high degree of mechanisation).

FULCO Implemented South Deep commenced Full Calendar Operations (“FULCO”) in February 2010.

• FULCO is described by GFI as a 6-day “long-off” working cycle. It increases the number of employees (and overall number of shifts worked) but reduces individuals’ shifts; employees generally work a six-day week but have a four-day break every fourth week. Maintenance is incorporated into the FULCO cycle.

• GFI believes the FULCO system will yield various benefits, e.g.: better productivity, improved employee morale. • The cost to GFI is a wage premium of 10%, but GFI believes that production would increase 30%-35%.

We have our doubts as to whether FULCO will work, given past failures at other companies. However, GFI believes even in the event that FULCO fails, South Deep would be no worse off than before. In our view, the key to FULCO is to ensure that head grades are maintained or improved, in which case FULCO would work very well. However, if tonnages are increased at the expense of grade, FULCO could prove disastrous – not only would GFI have to try to shed labour, but wage rates tend to be “sticky” (easy to move up, but not down).

South Shaft – Beyond 330 ktpm South Shaft is linked to the Twin Shafts at 95 level, and is currently contributing 60 ktpm of hoisting capacity. Infrastructure work is taking place that will see capacity increase to 120 ktpm over 2-2.5 years.

• South Shaft has a 1 Moz reserve that was originally planned to be extracted in later years, but GFI now plans to bring South Shaft into production by July 2010. Initial production rates will be around 20 ktpm @ 4.5-6 g/t, at a capex cost of R55 million and opex of R380/t.

• Ultimately, South Shaft could be contributing production of 150 ktpm by the time the Twin Shafts hit 330 ktpm. Planning is currently being undertaken to expand the South Deep plant accordingly.

Potential For Future Synergies With Kloof At some point in the future, South Deep will undergo a Phase 2 expansion.

• This will involve driving a decline from 110 level at the bottom of South Deep’s Main Shaft to 135 level (around 3.5 km below surface), in the direction of Kloof 4 Shaft.

• The 45 level at Kloof’s 4 Shaft is already at a depth equivalent to South Deep’s 135 level; the two are separated by around 4 km. • Connecting the two would allow around 150 ktpm to be hoisted via Kloof.

Key Milestones To Watch For • Full-year FY10 production of 300 koz (currently on-track). • The Vent Shaft is to be commissioned by July 2012. • Full production (330ktpm, or 750-800 koz annually) is scheduled for December 2014. • South Shaft should commence some production in FY10-FY11 and be producing at 120 ktpm by FY12-FY13.

We emphasise that South Deep’s 330 ktpm target excludes production from South Shaft and from VCR at the Twin Shafts. Actual production should therefore be just under the 450 ktpm mark if all goes to plan.


29

Gold One International

Modder East Gold Mine RBC visited Gold One’s (ASX & JSE: GDO) Modder East operation on Thursday, January 28th.

Notable Features Of Modder East • First New East Rand Mine - Modder East, located approximately 30km to the east of Johannesburg, was the first new mine on

the East Rand for 28 years. First gold was poured in July 2009 (240 oz), following plant commissioning in June 2009. • Targeting BPLZ - Modder East’s principal target is the 0.5m-wide high-grade Buckshot Pyrite Leader Zone (“BPLZ”; dips

approximately 3°-6° south). The BPLZ is the topmost of three facies that make up the Black Reef. It overlies the 1.3m-thick low-grade Blanket Facies, which in turn sits above the 3m-wide low-grade Channel Facies.

• Shallow - Mining is shallow (300m-500m), using conventional breast stoping. Stope widths are around 1m-1.8m, depending on reef thickness (or more accurately, depending on grade in the underlying facies).

• Mechanised - Mechanised decline and footwall development means more efficiency and less dilution than more-traditional reef drives. Truck hoisting saves electricity; unlike at most other mines, the (345m) vertical shaft will be used principally to move men, and for ventilation.

• Hydropower Drilling - Hydropower drilling also saves on electricity and is helping keep opex low (hydropower drilling is 30% cheaper than compressed air drilling; capex is 40% cheaper) and has productivity benefits (e.g., drilling is twice as quick as pneumatic drills, so less rock drill operators would be required).

Exhibit 18: Mining BPLZ. Gold mineralization is hosted within the buckshot layers.


Ramp-Up Going Well • Commercial Production Declared, Averts Bond Reset - “Commercial and continuous” production was declared in December

2009. This beat the March 2010 deadline that had been a condition of GDO’s US$71.6 million convertible bonds. Missing the deadline would have resulted in a reset of the bonds’ conversion price in the favour of the bondholders.

• 2009 Production In-line - GDO had targeted production of 16-20 koz in 2009 and actually achieved 17 koz, with the December quarter contributing 10.9 koz. Cash cost for the one month of commercial production in December averaged $593/oz, versus guidance of $640/oz.

• 2010 Target Should Be Achievable - December’s 5.2 koz annualized is 62 koz, so 2010’s target rate of 100-120 koz (at US$330/oz cash cost) should be within reasonably-easy reach. The 2011 (steady-state) target is 180 koz at US$250/oz.

• Tonnages and Flexibility Should Improve - Average face advance in the December quarter was 12.8m/month, just shy of the targeted 14m/month. There will be significantly more reef attack points going forward (e.g., 8 main ends in 2010) than was enjoyed in 2009 (2 main ends), giving GDO more production flexibility.

• Plant to Expand - The plant has been commissioned with an initial 70 ktpm capacity; a secondary crushing circuit is being installed to increase this to 100 ktpm. Meanwhile, testwork has shown that a gravity circuit should improve recoveries and reduce opex and gold lock-up. Consequently, a gravity circuit is to be completed by May 2010.

• Capex Well-Controlled - Construction capex was $108 million (R814 million), sustaining capex is expected to be $100/oz of production. EPCM was done mainly in-house and commendably came in R100 million under-budget. This is despite that the plant and mining equipment are all brand new.

• Excellent Recoveries, But Should Fall Going Forward - Recoveries are currently in the order of 92%, versus a BFS value of 87% (95% for UK9A); however, we caution that this is being achieved on lower volumes (25 ktpm) and recoveries could fall as volumes rise (to 75 ktpm).


30

Exhibit 19: Conventional Plant. Modder East’s plant essentially comprises a semi-autogenous mill, CIL circuit, and electrowinning and smelting facilities. The oven shown below is in the plant’s laboratories, and is not the smelter.


Potential For More At Modder East • Longer Life Likely - LOM should be longer than official 8 years for “Phase 1”, as GDO will come back to mine Blanket and

Channel facies (only a small amount of Blanket and no Channel is included in the current resource). Strike gullies are being dug deeper than necessary to better evaluate these facies; where there are thick areas (up to 4m) of high grade, bulk mining is a possibility. It should also be noted that gullies and footwall contain at least some grade, whereas these were assigned zero grade in feasibility studies.

• Current Resource Declared At Lower Prices - GDO’s 3.7 Moz resource (2.63 Moz Indicated, 1.04 Moz Inferred) was declared at $658/oz and R6.29/USD, so the resource should be much larger at today’s gold price in any case.

• Kimberley Upside Potential - Kimberley Reef represents only 300 koz of the current reserve; there is thus potential for upside in the Kimberleys (which sit at 290m-500m and dip at up to 15°-20° south). On current projections, at steady state, around 70% of production should come from Black Reef (mostly BPLZ facies), and the remainder from Kimberley UK9a reef.

• Kimberley Access Soon - Because of the patchy nature of Kimberley mineralisation, GDO intends to conduct further exploration from underground via exploration drives. The decline is 100m away from where it will branch off to access the Kimberley reef; GDO expects Kimberley Reef access around March-May of 2011.


31

Miscellaneous Issues At Modder East/East Ramd • New Order Mining Right Expected Soon – The Old Order right conversion application was submitted a year before deadline;

New Order right expected in next few months. • Largely Unflooded - None of GDO’s key mining areas are associated with any flooding. • Power Unlikely To Be A Problem - GDO has never experienced load-shedding and don’t expect to in future, as nearby refineries

and pumping at the neighbouring Grootvlei property cannot be shut down. • Electricity Is Small Proportion Of Costs - Electricity is currently 7% of cash costs; three successive 35% annual tariff increases

would bring this to 9% in 2012.

Ventersburg - Next Flagship? The Ventersburg project is located in the Free State, approximately 25 km southeast of Welkom. • Shallow A-Reef - The project is targeting A Reef, which occurs as a 2 km-wide channel on the property, at depths of 465m-850m;

i.e., relatively-shallow. • Resource Addition Expected Soon - The current resource is 1.44 Moz @ 5.1 g/t Indicated and 1.84 Moz @ 4.24 g/t Inferred.

Second phase drilling commenced July 2009, and a “substantial” resource addition might occur in early 2010. • GFI Participation Right - GFI has participation right for up to 51%, albeit limited to Area 1 (Areas 2 and 3 have potential). • Timeline - A PFS is being reviewed, and a BFS is expected in 2011. If all goes to plan, development will occur over 2012-2014,

with production to commence in Q4-2014. • Bigger But Lower-Grade Than Modder East - Ventersburg should be bigger (volume-wise) but lower-grade than Modder East.

It would comprise deep shafts with a combination of mechanised and conventional mining. Capex is estimated at US$200-300 million.

East Rand Boundary Projects (“ERBP”) The ERBP comprises Holfontein, Turnbridge and the northern part of Modder East. • Some Production Potential - GDO believes there is potential for 50 kg/month from unflooded areas; ore would be delivered to the

Modder East plant (which will be fine-tuned to add extra capacity). • Production Could Start In 2010? - GDO was projecting a two-year lead time and 5-year life-of-mine, although management

recently said that ERBP could be producing as early as 2010.

Vlakfontein - Probable Spin-Off • Deep Project - Vlakfontein comprises two deep shafts, intersecting Kimberley Reef around 600m and Main Reef at 2 km. • Spin-Off Candidate - This contradicts GDO’s positioning as a “shallow, low-technical-risk producer”. Consequently, we believe

Vlakfontein is likely to be spun off.

Tulo • Lake Is Key - Tulo is located in a remote area of Mozambique, 170 km north of Lichinga, close to the shores of Lake Malawi.

River flooding limits road access to around 6 months per year. However, GDO views the lake as potentially providing year-round access to Tulo, and thereby a key to unlocking the project’s potential.

• Bulk Sampling In 2010? - There is currently a 32 koz alluvial resource that could potentially be mined in 2010 as a “bulk sampling exercise”.

• Find The Vein(s)! - GDO believes that the gold in this resource originated from shear-hosted veins, and it is these veins that would be the real prize at Tulo. To this end, an airborne magnetic survey and exploration drilling are to commence in Q2-10 and 2011, respectively.

Corporate Strategy • 500 koz Target, But No Rush - GDO has set itself an internal target of producing 500 koz in around 5 years. To this end, GDO

will consider acquisitions for growth (beyond Modder East, ERBP and Ventersdorp) when share price is “better reflective” of what management believes company is worth. Management believes GDO is currently undermarketed, and does not wish to use “undervalued” equity to purchase “overvalued” companies.

• Another Listing Soon? - GDO’s shares trade far more actively on the ASX than on the JSE; the percentage of ASX-listed shares has increased from 5% to >50%. We believe GDO will be looking to list in North America, in the hopes of further improving visibility and liquidity.


32

• Debt Facility Being Arranged - GDO’s convertible bondholders have a put option exercisable on December 12th, 2010. Although GDO was already cash-flow positive in December 2009 (including capital - but not bond interest), we do not expect GDO to have generated sufficient cash in the unlikely event that all (or a substantial portion) of bondholders elect to exercise. Prudently, GDO is looking to arrange a debt facility to cover any funding requirements that should arise as a result.

Favourable Outlook Things appear to be going swimmingly for GDO at the moment. • Modder East is ramping up production at a time of high gold prices, and possesses potential production and life-of-mine upside. • A second mine is being lined up in the form of Ventersdorp; there are a number of other projects at various stages in the pipeline. • On the whole, these projects are relatively-shallow and relatively-low technical-risk.


33

Great Basin Gold

Burnstone Gold Mine RBC visited Great Basin Gold’s (TSX: GBG; “Great Basin”) Burnstone project on Friday, January 29th.

Burnstone is located around 80 km southeast of Johannesburg and just east of Balfour. It is expected to produce 254 koz annually over a 19-year mine life, for a capital cost of US$224 million and at a cash cost of around US$319/oz.

Exhibit 20: Longhole Stoping – Narrow Kimberley Reef (Middle Row) and Wide Reef (Bottom Row).

Source: RBC Capital Markets, Great Basin Gold

Longhole Stoping Trials Were Proceeding Well Development is underway at B (already on-reef) and C blocks; longhole stoping trials are already in progress at B block. • Key success factors to longhole stoping are consistently-accurate drilling and appropriate blasting forces. • RBC was able to view two trial stopes (one “narrow reef” and one “wide reef”), both of which appeared to have been successfully

mined. • On questioning, RBC learnt that resue mining was being considered for development (to reduce dilution of development ore).

Other Visit Observations • The vertical shaft has nearly reached bottom (around 40m to go). Shaft commissioning should occur around June 2010. • The decline has reached a depth of 350m; it is around 234m from intersecting the vertical shaft (intersection should occur in April

2010).


34

• The plant site has been prepared; plant construction proper should commence in February. Mill refurbishment was reportedly 88% complete.

• The permanent winder house is already being built and should be completed in February 2010; commissioning of the refurbished winders should occur in April 2010.

• The waste rock dump should be completed in May 2009; GBG is already (temporarily) loading a portion with 60 kt of low-grade development ore (ready for plant commissioning).

Power Coming Great Basin currently has 2 MVA of grid power; it also has 12 MVA of generators on-site. GBG expects to receive 25 MVA of grid power from Q2-CY10, and 52 MVA from Q3-CY10 (60 MVA capacity).

Company Expectations • First gold pour at Burnstone is scheduled for July 2010. This may or may not prove over-optimistic – while we view GBG’s

proposed timeline as tight-but-not-unachievable, we note that projects such as this have a tendency to slip by up to a few months. • Capex will be $118 million for 2010, and $52 million in 2011. • LOM average cash cost is expected to be $392/oz, with power comprising $25/oz or 7%. Should there be three years of 35% tariff

increases, the proportion of costs attributable to power could rise to 11%.

Lower Opex If Longhole Stoping Successfully Adopted There is potential for still lower unit cash costs (15% lower; from lower dilution and processing efficiencies) if longhole stoping is adopted successfully.

• Capex would be higher, however, as would wages (fewer workers, but they would be more highly-skilled and highly-paid). • Consequently, total cost/tonne would be similar for both methods. • However, longhole stoping has another key benefit: vastly-improved safety (and therefore lower likelihood of economically-

damaging stoppages).

Exhibit 21: Winder House and Plant Site Being Prepared. Burnstone’s plant will be a conventional one, essentially comprising semi-autogenous and ball mills, CIL circuit, and electrowinning and smelting facilities.

Source: RBC Capital Markets, Great Basin Gold

Conclusion – Longhole Trials Looking Good • If longhole stoping trials continue to be as successful as has been the case thus far, we believe the method will almost certainly be

adopted. • The reduced dilution would allow more gold to be produced assuming the same tonnage is put through the plant. • Great Basin is now set to deliver two mines in close succession, and there are short-term intermediate catalysts aplenty.


35

Pan African Resources

Barberton Gold Mine On January 25th, RBC visited the 62-11 block at the Fairview Section of Pan African Resource’s (AIM: PAF; “Pan African”) Barberton Mine (the other two sections being Sheba and New Consort), some 500m below sea-level (around 1.4 km below surface). Access was via an adit from surface, followed by a series of chairlifts to 42 Level, a rail-bound cage (“giraffe”) in a 45.5°-dipping incline shaft to 62 Level and then a shallow decline to 62-11.

An Atypical South African Gold Mine • Unlike most South African gold mines, Barberton does not mine the reefs of the Witswatersrand Basin; instead, Barberton sits

some 370km east of Johannesburg and 47km southwest of Nelspruit, where it mines the Barberton Greenstone Belt (“BGB”). • Fairview Section is situated along the central and southern portions of two synclines, Eureka and Ulundi. Most of Fairview’s

production comes from the Sheba Fault Zone between the two synclines. • Barberton’s orebodies are refractory; much of Barberton’s gold production uses a bacterial oxidation process, BIOX® (the BIOX®

process was in fact developed originally at Fairview).

Deceptively-Small Reserves and Resources – Barberton Is The AA-Sized Duracell Of Mines • Current reserves are estimated at 6 years’ (0.6 Moz); resources, at 20 years’ (2 Moz). The on-going reserve replacement

programme is aimed at increasing these to 8 and 30 years, respectively. • That said, Barberton’s life-of-mine has typically been estimated at around 5-to-10 years since 1885 or so. However, the continual

discovery of new orebodies, a trend we expect is likely to continue, has allowed the mine to continue operating to present day. • At Sheba, a decline is being driven towards the Thomas and Joe’s Luck orebodies from the Edwin Bray development; there is

potential for more orebodies to be discovered along the way. • Pan African possesses a New Order prospecting right over a large, largely-unexplored and highly-prospective anticlinal structure to

the south of the present mining areas. However, a dispute with park authorities has to be resolved before any work can take place.

Orebody Suits Mechanisation The steep-dipping nature of Barberton’s orebodies facilitates the use of serni-mechanised cut-and-fill as the dominant mining method, resulting in higher productivity and lower costs. • Cut-and-fill allows veins to be followed more flexibly, improves sidewall stability and reduces waste dilution. • Essentially, 2m cuts are taken from the hanging wall (i.e., the stope back). Waste rock capped with cemented backfill serves as the

working platform for mining further cuts above, thus reducing the amount of waste that has to be hoisted out of the mine. At Fairview, LHDs load ore onto conveyor belts via ore passes, and the ore is subsequently transferred to the incline rail-bound conveyances for transport to surface.

• In-situ grades can be as high as 100 g/t, although the average reserve and resources grades are just over 8 g/t. Pan African is targeting the maintenance of a mill headgrade of >10 g/t.

Fairview and Sheba each account for around 40% of Barberton’s mined tonnage (a combined 30 ktpm or so, or 90-100 koz annually), New Consort for the remainder.


36

Exhibit 22: Chairlift And Giraffe (Yellow Box) To Cut-and-Fill Stope.


(Almost) Unique Plant Much of Barberton’s resources are refractory in that gold is often occluded within pyrite and arsenopyrite crystals. There are essentially two means of liberating gold from these sulphide minerals: roasting and biological oxidation (BIOX®). Fairview is one of a handful plants globally to be using BIOX®. • BIOX® Advantages Over Roasting – These potentially include higher recoveries, no environmentally-unfriendly sulphur dioxide

emissions (saving on acid plant capex), lower opex, ability to operate at lower concentrate sulphur grades (reducing flotation costs and losses to tailings).

• Arsenic No Problem - The presence of arsenopyrites means that arsenic comprises around 4% of the flotation concentrates. The BIOX® effluent contains ferric sulphate; increasing the pH results in the precipitation of highly-stable ferric arsenate, which can be safely deposited into slimes dams.


37

Run-of-Mine (“ROM”) ore is treated at concentrators located at the respective sections; flotation concentrates are further treated at the BIOX® plant. Gravity and BIOX® concentrates are smelted, and are sent for refining at the Rand Refinery.

Recovery at the flotation plants is around 92%-95%; the BIOX® plant at Fairview recovers around 97% from the flotation concentrate fed it. Overall recovery from all operations is around 91%-92%.

Exhibit 23: Ore Is Transported From Fairview’s Shaft Head To The Plant By Cable Car.

Source: RBC Capital Markets, Pan African Resources

Capex & Opex • Cash cost in H1-FY10 was R165,000/kg, or around US$670/oz. However, costs were adversely affected by a two-week stoppage

in December 2009 to root out criminal miners. Pan African estimated that had not this operation been undertaken, costs would have been around 152,000/kg or US$620/oz.

• H1-FY10 capex at Barberton was £2.2 million (£1.27 million development capex, £0.933 million maintenance capex).

Outlook for Barberton: Improved Performance Expected • Production should increase in H2-FY10 and costs should fall on the back of higher volumes. • At current gold prices, Barberton’s cash flows should be sufficient to fund on-going capex requirements, existing growth projects

and exploration programmes. • There remains exploration potential aplenty, despite the operation’s long production history.


38

Companies Mentioned

Aquarius Platinum (ASX & LSE: AQP - £3.57; Outperform, Above Average Risk)

Eastern Platinum (TSX & AIM: ELR - C$1.35; Outperform, Average Risk)

Jubilee Platinum (AIM: JLP - £0.34; Outperform, Speculative Risk)

Platinum Australia (ASX: PLA; AIM: PLAA - A$1.04; Outperform, Above Average Risk)

Platmin (TSX & AIM: PPN – C$1.33; Outperform, Above Average Risk)

Sylvania Resources (AIM: SLV - £0.53; Sector Perform, Above Average Risk)

Central Rand Gold (LSE: CRND - £0.15; Outperform, Speculative Risk)

DRDGOLD (JSE: DRD – R4.90; Sector Perform, Average Risk)

Gold Fields (JSE: GFI – R93.85; Sector Perform, Average Risk)

Gold One International (JSE & ASX: GDO – R2.09; Outperform, Above Average Risk)

Great Basin Gold (TSX: GBG – C$1.76; Outperform, Average Risk)

Pan African Resources (AIM: PAF - £0.06; Outperform, Speculative Risk)

Impala Platinum (JSE: IMP; R196.20; Sector Perform, Above Average Risk)

Lonmin (LSE: LMI; £18.71; Outperform, Average Risk)

Nkwe Platinum (ASX: NKP; A$0.50; Sector Perform, Speculative Risk)

Northam Platinum (JSE: NHM; R49.00; Outperform, Above Average Risk)

Priced as of market close February 17, 2010 ET.


39

Required Disclosures

Non-U.S. Analyst Disclosure

Yuen Low, Leon Esterhuizen, and Arnold van Graan (i) are not registered/qualified as research analysts with the NYSE and/or FINRAand (ii) may not be associated persons of the RBC Capital Markets Corporation and therefore may not be subject to FINRA Rule 2711and NYSE Rule 472 restrictions on communications with a subject company, public appearances and trading securities held by aresearch analyst account.

Conflicts Disclosures

This product constitutes a compendium report (covers six or more subject companies). As such, RBC Capital Markets chooses toprovide specific disclosures for the subject companies by reference. To access current disclosures for the subject companies, clientsshould refer to http://www7.rbccm.com/GLDisclosure/PublicWeb/DisclosureLookup.aspx?EntityID=1 or send a request to RBC CMResearch Publishing, P.O. Box 50, 200 Bay Street, Royal Bank Plaza, 29th Floor, South Tower, Toronto, Ontario M5J 2W7.

The analyst(s) responsible for preparing this research report received compensation that is based upon various factors, including totalrevenues of the member companies of RBC Capital Markets and its affiliates, a portion of which are or have been generated byinvestment banking activities of the member companies of RBC Capital Markets and its affiliates.

Distribution of Ratings

For the purpose of ratings distributions, regulatory rules require member firms to assign ratings to one of three rating categories - Buy,Hold/Neutral, or Sell - regardless of a firm's own rating categories. Although RBC Capital Markets' ratings of Top Pick/Outperform,Sector Perform and Underperform most closely correspond to Buy, Hold/Neutral and Sell, respectively, the meanings are not the samebecause our ratings are determined on a relative basis (as described above).

Distribution of RatingsRBC Capital Markets, Equity Research

Investment BankingServ./Past 12 Mos.

Rating Count Percent Count Percent

BUY[TP/O] 590 49.60 180 30.51HOLD[SP] 533 44.80 123 23.08SELL[U] 67 5.60 9 13.43

Conflicts Policy

RBC Capital Markets Policy for Managing Conflicts of Interest in Relation to Investment Research is available from us on request. Toaccess our current policy, clients should refer tohttps://www.rbccm.com/global/file-414164.pdfor send a request to RBC CM Research Publishing, P.O. Box 50, 200 Bay Street, Royal Bank Plaza, 29th Floor, South Tower,Toronto, Ontario M5J 2W7. We reserve the right to amend or supplement this policy at any time.

Dissemination of Research and Short-Term Trading Calls

RBC Capital Markets endeavours to make all reasonable efforts to provide research simultaneously to all eligible clients, havingregard to local time zones in overseas jurisdictions. RBC Capital Markets' research is posted to our proprietary websites to ensureeligible clients receive coverage initiations and changes in rating, targets and opinions in a timely manner. Additional distribution maybe done by the sales personnel via email, fax or regular mail. Clients may also receive our research via third party vendors. Pleasecontact your investment advisor or institutional salesperson for more information regarding RBC Capital Markets research.RBC Capital Markets also provides eligible clients with access to a database which may contain Short-Term trading calls on certain ofthe subject companies for which it currently provides equity research coverage. The database may be accessed via the followinghyperlink https://www2.rbccm.com/cmonline/index.html. The information regarding Short-Term trading calls accessible through thedatabase does not constitute a research report. These Short-Term trading calls are not formal ratings and reflect the research analyst'sviews with respect to market and trading events in the coming days or weeks and, as such, may differ from the price targets andrecommendations in our published research reports reflecting the research analyst's views of the longer-term (one year) prospects ofthe subject company. Thus, it is possible that a subject company's common equity that is considered a long-term 'sector perform' oreven an 'underperform' might be a Short-Term buying opportunity as a result of temporary selling pressure in the market; conversely,a subject company's common equity rated a long-term 'outperform' could be considered susceptible to a Short-Term downward pricecorrection.


http://www7.rbccm.com/GLDisclosure/PublicWeb/DisclosureLookup.aspx?EntityID=1

https://www.rbccm.com/global/file-414164.pdf

https://www2.rbccm.com/cmonline/index.html

40

Analyst Certification

All of the views expressed in this report accurately reflect the personal views of the responsible analyst(s) about any and all of thesubject securities or issuers. No part of the compensation of the responsible analyst(s) named herein is, or will be, directly orindirectly, related to the specific recommendations or views expressed by the responsible analyst(s) in this report.

Disclaimer

RBC Capital Markets is the business name used by certain subsidiaries of Royal Bank of Canada, including RBC Dominion Securities Inc., RBC Capital MarketsCorporation, Royal Bank of Canada Europe Limited and Royal Bank of Canada - Sydney Branch. The information contained in this report has been compiled by RBCCapital Markets from sources believed to be reliable, but no representation or warranty, express or implied, is made by Royal Bank of Canada, RBC Capital Markets, itsaffiliates or any other person as to its accuracy, completeness or correctness. All opinions and estimates contained in this report constitute RBC Capital Markets'judgement as of the date of this report, are subject to change without notice and are provided in good faith but without legal responsibility. Nothing in this reportconstitutes legal, accounting or tax advice or individually tailored investment advice. This material is prepared for general circulation to clients and has been preparedwithout regard to the individual financial circumstances and objectives of persons who receive it. The investments or services contained in this report may not besuitable for you and it is recommended that you consult an independent investment advisor if you are in doubt about the suitability of such investments or services. Thisreport is not an offer to sell or a solicitation of an offer to buy any securities. Past performance is not a guide to future performance, future returns are not guaranteed,and a loss of original capital may occur. RBC Capital Markets research analyst compensation is based in part on the overall profitability of RBC Capital Markets, whichincludes profits attributable to investment banking revenues. Every province in Canada, state in the U.S., and most countries throughout the world have their own lawsregulating the types of securities and other investment products which may be offered to their residents, as well as the process for doing so. As a result, the securitiesdiscussed in this report may not be eligible for sale in some jurisdictions. This report is not, and under no circumstances should be construed as, a solicitation to act assecurities broker or dealer in any jurisdiction by any person or company that is not legally permitted to carry on the business of a securities broker or dealer in thatjurisdiction. To the full extent permitted by law neither RBC Capital Markets nor any of its affiliates, nor any other person, accepts any liability whatsoever for anydirect or consequential loss arising from any use of this report or the information contained herein. No matter contained in this document may be reproduced or copiedby any means without the prior consent of RBC Capital Markets.

Additional information is available on request.

To U.S. Residents:This publication has been approved by RBC Capital Markets Corporation (member FINRA, NYSE), which is a U.S. registered broker-dealer and which acceptsresponsibility for this report and its dissemination in the United States. Any U.S. recipient of this report that is not a registered broker-dealer or a bank acting in a brokeror dealer capacity and that wishes further information regarding, or to effect any transaction in, any of the securities discussed in this report, should contact and placeorders with RBC Capital Markets Corporation.To Canadian Residents:This publication has been approved by RBC Dominion Securities Inc.(member IIROC). Any Canadian recipient of this report that is not a Designated Institution inOntario, an Accredited Investor in British Columbia or Alberta or a Sophisticated Purchaser in Quebec (or similar permitted purchaser in any other province) and thatwishes further information regarding, or to effect any transaction in, any of the securities discussed in this report should contact and place orders with RBC DominionSecurities Inc., which, without in any way limiting the foregoing, accepts responsibility for this report and its dissemination in Canada.To U.K. Residents:This publication has been approved by Royal Bank of Canada Europe Limited ('RBCEL') which is authorized and regulated by Financial ServicesAuthority ('FSA'), inconnection with its distribution in the United Kingdom. This material is not for general distribution in the United Kingdom to retail clients, as defined under the rules ofthe FSA. However, targeted distribution may be made to selected retail clients of RBC and its affiliates. RBCEL accepts responsibility for this report and itsdissemination in the United Kingdom.To Persons Receiving This Advice in Australia:This material has been distributed in Australia by Royal Bank of Canada - Sydney Branch (ABN 86 076 940 880, AFSL No. 246521). This material has been preparedfor general circulation and does not take into account the objectives, financial situation or needs of any recipient. Accordingly, any recipient should, before acting onthis material, consider the appropriateness of this material having regard to their objectives, financial situation and needs. If this material relates to the acquisition orpossible acquisition of a particular financial product, a recipient in Australia should obtain any relevant disclosure document prepared in respect of that product andconsider that document before making any decision about whether to acquire the product.To Hong Kong Residents:This publication is distributed in Hong Kong by RBC Investment Services (Asia) Limited and RBC Investment Management (Asia) Limited, licensed corporationsunder the Securities and Futures Ordinance or, by Royal Bank of Canada, Hong Kong Branch, a registered institution under the Securities and Futures Ordinance. Thismaterial has been prepared for general circulation and does not take into account the objectives, financial situation, or needs of any recipient. Hong Kong personswishing to obtain further information on any of the securities mentioned in this publication should contact RBC Investment Services (Asia) Limited, RBC InvestmentManagement (Asia) Limited or Royal Bank of Canada, Hong Kong Branch at 17/Floor, Cheung Kong Center, 2 Queen's Road Central, Hong Kong (telephone numberis 2848-1388).To Singapore Residents:This publication is distributed in Singapore by RBC (Singapore Branch) and RBC (Asia) Limited, registered entities granted offshore bank status by the MonetaryAuthority of Singapore. This material has been prepared for general circulation and does not take into account the objectives, financial situation, or needs of anyrecipient. You are advised to seek independent advice from a financial adviser before purchasing any product. If you do not obtain independent advice, you shouldconsider whether the product is suitable for you. Past performance is not indicative of future performance.

®Registered trademark of Royal Bank of Canada. RBC Capital Markets is a trademark of Royal Bank of Canada. Used under license.Copyright © RBC Capital Markets Corporation 2010 - Member SIPC

Copyright © RBC Dominion Securities Inc. 2010 - Member CIPFCopyright © Royal Bank of Canada Europe Limited 2010

Copyright © Royal Bank of Canada 2010All rights reserved


Documents

Royal Bank of Canada Europe Limited INDUSTRY | … notes/2010...Should this be the case, we would expect a capital raise in excess of US$100 million. Jubilee Platinum (“Jubilee”)