RDS in Brazil CONNECTIONS SERIES. Longer-term gain

Transcription

1 Global/Europe/Latin America Equity Research Oil & Gas The Credit Suisse Connections Series leverages our exceptional breadth of macro and micro research to deliver incisive cross-sector and cross-border thematic insights for our clients. Research Analysts Thomas Adolff Andre Natal Regis Cardoso Specialist Sales: Jason Turner RDS in Brazil CONNECTIONS SERIES Longer-term gain Ahead of RDS' field trip to Brazil in early November, this report provides a preview and analysis of RDS' assets in Brazil, with a focus on its upstream assets in the Santos Basin. Bottom line, we estimate RDS' assets there have the potential to grow production from ~240kbd in 2016E to ~800kbd by 2025E (through a 'prioritised' development of Libra in the 2020s), driving CFFO from ~$1.5bn (at ~$45/bbl Brent) in 2016E to ~$9bn (at $70/bbl Brent) by 2025E. FCF over E could average ~$7.5bn pa with further upside potential. Reform should be forthcoming. The Brazilian government is sending positive signals on potential reforms a more pragmatic approach may be forthcoming to the benefit of the entire oil and gas industry in Brazil. Momentum in efficiency gains. These include continuing declines in drilling and completion (D&C) days, faster ramp-up to plateau and longer plateau duration than budgeted. FCF potential may surprise over time. Importantly, the consortium is successfully mitigating local content challenges related to the locally-built production units, thereby continuing to de-risk the outlook against its 'latest' guidance. Long(er) lived, yet conventional in nature. Data to date have exceeded expectations with constant production, good behaviour of the reservoirs, good lateral communication and no significant issues regarding flow guarantee. Plateau duration should exceed the current base case. Figure 1: RDS Brazil Upstream Free Cash Flow profile ($mn) $10,000 $8,000 $6,000 $4,000 $2,000 $0 -$2, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E FCF FCF (BG assets only) Source: Credit Suisse estimates; Note: Brent ~$54/bbl (2015), ~$44 (2016), ~$56 (2017), ~$68 ( ) and $70 (2020+) DISCLOSURE APPENDIX AT THE BACK OF THIS REPORT CONTAINS IMPORTANT DISCLOSURES, ANALYST CERTIFICATIONS, LEGAL ENTITY DISCLOSURE AND THE STATUS OF NON-US ANALYSTS. US Disclosure: Credit Suisse does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the Firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision.

2 Table of contents Key charts at CS macro 3 Key charts at $50/bbl flat 4 Executive summary 5 RDS in Brazil post the BG deal 6 Long friendship, but only now material in nature...6 Deep Water Importance of Brazil to RDS...7 Development strategy phased and standardised...8 Developments in the pre-salt...9 Mitigating operational delays...12 Features of Long-life / Long plateau...12 Improving recovery factor...14 Ramp up period...17 Well performance...18 Drilling time...19 Gas monetization...19 Unitisation challenges...20 License period...23 Libra...24 Access to more resources...25 Exploration in Brazil...25 Reforms...25 Key Upstream Asset description 28 Lula/Iracema (BM-S-11)...28 Greater Iara Area (BM-S-11)...29 Libra...31 Sapinhoa (BM-S-9)...32 Lapa (BM-S-9)...32 RDS in Brazil 2

3 Key charts at CS macro Figure 2: RDS (+BG) Upstream Brazil FCF ($mn) $10,000 $8,000 $6,000 $4,000 Figure 3: BG Upstream Brazil only FCF ($mn) $10,000 $8,000 $6,000 $4,000 $2,000 $2,000 $0 $0 -$2,000 -$4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E -$2,000 -$4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E CFFO CAPEX FCF CFFO CAPEX FCF Source: Credit Suisse estimates Source: Credit Suisse estimates Figure 4: RDS FCF ($mn) vs entitl production (kboed) Figure 5: BG only FCF ($mn) vs entitl production (kboed) $10, $10, $8,000 $6, $8,000 $6, $4, $4, $2, $2, $0 -$2, $0 -$2, $4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 0 -$4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 0 FCF Production, rhs FCF Production, rhs Source: Credit Suisse estimates Source: Credit Suisse estimates Figure 6: RDS entitlement production profile (kboed) Figure 7: BG only production profile (kboed) E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E Lula/Iracema Sapinhoa Greater Iara area Lapa Libra BC-10 Lula/Iracema Sapinhoa Greater Iara area Lapa Source: Credit Suisse estimates; Note: we assume additional FPSOs than currently guided to in the 2020s, most notably for the Greater Iara area. Source: Credit Suisse estimates; Note: we assume additional FPSOs than currently guided to in the 2020s, most notably for the Greater Iara area. RDS in Brazil 3

4 Key charts at $50/bbl flat Figure 8: RDS (+ BG) Upstream Brazil FCF ($mn) $10,000 $8,000 $6,000 $4,000 Figure 9: BG Upstream Brazil only FCF ($mn) $10,000 $8,000 $6,000 $4,000 $2,000 $2,000 $0 $0 -$2,000 -$4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E -$2,000 -$4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E CFFO CAPEX FCF CFFO CAPEX FCF Source: Credit Suisse estimates Source: Credit Suisse estimates Figure 10: RDS FCF ($mn) vs entitl production (kboed) Figure 11: BG only FCF ($mn) vs entitl production (kboed) $10, $10, $8,000 $6,000 $4,000 $2,000 $0 -$2, $8,000 $6,000 $4,000 $2,000 $0 -$2, $4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 0 -$4, E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 0 FCF Production, rhs FCF Production, rhs Source: Credit Suisse estimates Source: Credit Suisse estimates Figure 12: RDS entitlement production profile (kboed) Figure 13: BG only production profile (kboed) E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E Lula/Iracema Sapinhoa Greater Iara area Lapa Libra BC-10 Lula/Iracema Sapinhoa Greater Iara area Lapa Source: Credit Suisse estimates; Note: we assume additional FPSOs than currently guided to in the 2020s, most notably for the Greater Iara area. Source: Credit Suisse estimates; Note: we assume additional FPSOs than currently guided to in the 2020s, most notably for the Greater Iara area. RDS in Brazil 4

5 Executive summary Importance of Brazil. Ahead of RDS' field trip to Brazil in early November, this report looks into its assets in Brazil, with a particular focus on its upstream assets in the Santos Basin, and the recent updates as far as efficiency as well as development risks are concerned. The players in the Santos Basin have done a good job in developing the resources despite technical and local content challenges. The average cost of extraction is less than $8/boe, according to PBR, and is gradually decreasing as the average time to drill and complete a well has fallen by 71% between 2010 and Bottom line, we estimate RDS' Brazil assets can grow entitlement production from ~240kbd in 2016E to ~800kbd by 2025E (which assumes a 'prioritised' development of Libra in the 2020s), thereby driving CFFO from ~$1.5bn (at ~$45/bbl Brent) in 2016E to ~$9bn (at $70/bbl Brent) by 2025E. FCF over E could average ~$7.5bn pa a material portfolio. Further positive surprises? Production from pre-salt fields in early May 2016 have exceeded 1mbd (on a gross basis) for the first time 10 years since the first discovery and two years after reaching ~500kbd. This was achieved from 60 producers only (52 in the Santos Basin) with the average flow rate in the Santos basin at an impressive 25kbd, exceeding initial expectations of 15-20kbd, which at the time was seen already as an optimistic assumption. Based on July data, production from the pre-salt layer stood at 1.32mbd. It only took 40 wells to 'prove up' 30bn bbls of oil in the Brazilian pre-salt. For the GoM, it took 250, and for the North Sea, 900 to reach a similar milestone. As PBR stated, as you move from the Campos Basin south to the Santos Basin, everything gets bigger prospect sizes, water depth, permeability, porosity, distance to shore, CO2 content. What's more, there is further upside from improved recovery factor, which would be valuable. Long-lived despite conventional in nature. Results from EWTs (extended well tests) and performance from Lula Pilot, amongst other things, have exceeded expectations with constant production, good behaviour of the reservoirs, good lateral communication and no significant issues regarding flow guarantee to date. The planned FPSOs are positioned aerially over the discoveries and in view of better reservoir connectivity, continuity and deeper oil/water contact, the aerial coverage of FPSOs allows for a larger access to resources with the same number of wells (i.e., no incremental subsea/drilling capex). This allows for extended plateau production. For example, Galp now thinks that Lula Pilot could produce at plateau for 7 years or more having already produced at plateau for ~4 years. Mitigation measures. As far as operational risks from the 'Operation Car Wash' scandal are concerned, Galp stated that it is more confident now than this time last year on project delivery (when it pushed guidance to the right) as it is making good progress on mitigation measures (transferring work from Brazilian yards to Asian yards). Since that revised guidance, mitigating measures are de-risking the outlook for the replicant FPSOs, potentially rendering Galp's latest guidance as conservative. Against this guidance, on the 2Q16 conference call Galp stated that five of the six units are largely de-risked, whilst the last still faces potential issues with the hull; there is plenty of time to risk manage this. Reform should be forthcoming. The Brazilian Petroleum Institute (IBP), a non-profit private organization founded in 1957 that today comprises over 200 associated companies, is one of the most active and representative bodies of the oil & gas industry in Brazil. The IBP has been pushing for reforms in the oil sector on a number of fronts that would have long-term benefits for the sector, the country, and all the companies involved. The Brazilian government is sending positive signals as far as reform is concerned a more pragmatic approach may be forthcoming. This includes the potential removal of the mandatory PBR operatorship in the pre-salt, enhancement of local content policies by allowing for more flexibility and use of foreign suppliers, the simplification of the unitisation process, amongst other things. These reforms can come with many benefits, including the potential faster development of 'unitisable' areas. A licensing round for four unitisation areas may be forthcoming in 2017, including RDS' Gato do Mato, which had been on hold. RDS in Brazil 5

6 RDS in Brazil post the BG deal Long friendship, but only now material in nature RDS' presence in Brazil: RDS has operated in Brazil for over 100 years across both the downstream and upstream segments of the value chain; a truly Brazilian company, yet not Brazilian by origin. The company's 20% interest in the Libra pre-salt field in the Santos Basin is the crown in RDS' Brazilian portfolio, with an appraisal programme currently ongoing. It also operates several producing fields in the Campos basin, offshore Brazil, including the Bijupirá and Salema fields (RDS 80%), and the BC-10 field (RDS 50%), where in early 2016 it started production from Phase 3. The company used to have a stake in BM-S-8 it sold it before the ~1bn boe Carcara discovery was made, and STL recently agreed to acquire PBR's 66% stake in the block for $2.5bn. Exploration interests include a block in the São Francisco onshore basin area (RDS 60%; it used to have exposure to more licenses, but following poor results these have been relinquished), BM-S-54 in the offshore Santos Basin (RDS 80%) and BM-ES-27 in the Espirito Santo basin (RDS 17.5%). In downstream, it has a 18% interest in Brazil Companhia de Gas de São Paulo (Comgás), a natural gas distribution company and a 50% share in the Raízen joint venture with Cosan to produce ethanol (capacity of >2bn litres per year), sugar and electricity as well as supply, distribute and sell transport fuels, with a retail network of ~5500 sites. BG's Brazil assets: BG's main Brazilian assets consist of five offshore blocks in the Santos Basin, which consists of giant fields such as Lula (25%), Iracema (25%), Iara (25%) Saphinhoa (30%) and Lapa (30%). BG produced ~80 kbd in Brazil 2014, and we forecast this to increase to ~460kbd by 2020 (vs the latest target of 500kbd) as FPSOs are brought onstream. It also has a stake in the Sagitario discovery (BM-S-50) and operates 10 offshore exploration blocks in the Barreirinhas Basin, offshore northern Brazil. BG used to have a 60.1% stake in Comgas, which it sold for $1.7bn (plus $1bn in associated debt). This transaction was completed in November 2012 as part of BG's deleveraging plan. Under merger control rules implemented in 2012, the RDS-BG deal had to obtain regulatory clearance from the Administrative Council for Economic Defense (CADE), and the approval occurred earlier than we expected with no conditions attached. We did not expect regulatory opposition for the BG transaction. Capital gains tax (CGT) are applicable at the asset level in Brazil, but not necessarily at the corporate level. That is why no CGT was paid with RDS acquisition of BG, which completed in February Figure 14: RDS presence across the value chain Figure 15: IOC Leadership in deep-water Brazil Source: RDS presentation April 2015 Source: RDS presentation April 2015 RDS in Brazil 6

7 Deep Water Importance of Brazil to RDS RDS's deepwater portfolio is 'resourceful.' It produced ~450kbd from deep-water assets excluding BG's Brazil assets (around half of which from the US GoM) in 2015, but has a vast reserve and (2C) resource base of ~12bn boe. A big chunk of this comes from Nigeria, where there are certain above ground issues, but a relevant portion also comes from the US GoM. Despite being 'conventional' in nature, RDS's deepwater assets (both Nigeria and US GoM) could almost be viewed as 'long-lived'. The hub around Mars (US GoM) will grow from here and then has the potential to stay at stable levels for a long time given the vast resource base that can also tie back to these facilities. The same applies to Bonga (Nigeria), where the resource base can underpin growth/ stable production. The Brazil assets (currently producing >200kbd) that came with the BG transaction are the most cost competitive deepwater barrels globally, in our view. We think RDS paid closer to ~3.5bn boe (ie PBR's view on the reserve/resource base adjusted for BG's equity), which we think is fair at this stage of the upstream cycle for these assets, but was more conservative than BG's estimate of P of 4-6-8bn boe. The bulk of the difference in estimates between BG and PBR likely relates to Lula/Iracema and Iara. The difference in view relates mainly to (a) recovery factor, and partly to (b) in-place volumes. The PMean already assumes the lower end of the 3-17bps upside potential on recovery factor from the application of water-alternate-gas (WAG). Successful WAG injection could add 1.5bnboe to the 6bnboe mid-case, explaining a large part of BG s upside case to 8bnboe. Given the data to date, the reserve base has the potential to grow from PBR's base case. Figure 16: Pre-salt in Brazil overview of key blocks and companies involved Source: Credit Suisse Research based on PBR, ANP; Note: STL announced the acquisition of PBR's stake in Carcara, which is due to close during 3Q16 The players in the Santos Basin have done a good job in developing the resources despite technical challenges; challenges include gas rich, carbon dioxide contaminated carbonate reservoirs located beneath a ~2km salt canopy in ultra-deep waters. Despite these challenges, the average cost of extraction has totalled less than $8/boe, and is gradually decreasing as the average time to drill and complete a well has fallen by 71% between 2010 and Including other pre-salt discoveries, production from pre-salt fields in early RDS in Brazil 7

8 6 September 2016 May exceeded 1mbd (on a gross basis) for the first time; 10 years since the first discovery and two years after reaching ~500kbd. This was achieved from 60 producers only (52 in the Santos Basin) with the average flow rate in the Santos basin at an impressive 25kbd. It only took 40 wells to 'prove up' 30bn bbls of oil in the Brazilian pre-salt. For the GoM, it took 250, and for the North Sea, 900 to reach a similar milestone. As PBR stated as you move from the Campos Basin south to the Santos Basin, everything gets bigger prospect sizes, water depth, permeability, porosity, distance to shore, CO2 content. Figure 17: Geology, Challenges and Uncertainties Source: Credit Suisse Research Development strategy phased and standardised Ahead of the development of Brazil's pre-salt discoveries, PBR outlined a three-phased development strategy Phase 0, Phase 1A and Phase 1B. Phase 0 uses early production systems (EWTs, Pilots) on each field to enable extended testing of the reservoir's productivity and to analyse any operational issues there may be with producing from the field. As water depths have become progressively deeper, EWTs have also enabled new technology to be tested. In essence, this is an information gathering stage, which is important. Phase 1A exploits the resources using tried and tested technologies (eg FPSOs), which makes sense as this has the advantage of more timely development this is also being performed in a standardized manner, which is more cost efficient as well. Phase 1B is defined to continue Phase 1A, to optimise development plans (by that time, you have a much better understanding of the reservoir behaviour; production history etc) and to use new technologies tailored for pre-salt conditions. Lula Pilot The pilot came onstream in late 2010, and has been producing at plateau from fewer wells than previously planned for around four years now. This is the unit, where PBR is testing the reservoir model (including production from a horizontal well (P8H)). The horizontal well is the producer, which produced water (~1kbd) and was seen as a potential risk by the market, although we do not consider this a risk. RDS in Brazil 8

9 The argument that Lula Pilot has been producing without water and water only showing up after four years of production in early 2015 is generically correct, but does not explain why water has been observed in the P8H well brought onstream only in December Chemical analysis has confirmed that the water has come from the water injector, which is located in close proximity to the horizontal well, and that it is not related to formation water that can be observed after years of production. Other injectors are also close to other vertical producers (wells that can keep the unit at plateau), but the reason this well has produced water and others did not (aside from small amounts of condensed water, which any well would produce to the tune of barrels a day) is likely due to the proximity of the injector to producer pair and the layering of the reservoir/structure in this particular configuration. The fact that it is a horizontal well, so it had further reach, has possibly also contributed to the connectivity between wells. Indeed, comments by BG, Galp and PBR (including pre-salt managers) at the time suggested that no one had the slightest concern over this well (and that remains the case today), or the reservoir. Hence, there should not be an issue with the reserve and resource base, which over time has the potential to grow as recovery factors improve, which we discuss in a later section. Better understanding of reservoir behaviour (as encouraging as it might already be) is key and the pilot stages allow for testing and information gathering; over time, including success from WAG with CO2 floods, there might be further upside. Patience, however, is needed. Developments in the pre-salt Operationally, the development of the chartered FPSOs, where more work is being performed outside Brazil, are lower risk and the delivery has been more or less in line with the guided timeline (this is specific to the pre-salt Santos Basin, and not broadly applicable to Brazil's development track record). As far as the replicant (owned) FPSOs are concerned (see Figure 18), more operational uncertainty exist, as we highlighted in Sweet spots and the comeback of delays dated 30 September This is because more work is being performed domestically at shipyards in Brazil that lack both experience and financial strength. More recently, many of the local contractors got mired in the 'Operation Car Wash' scandal, which worsened the situation as far as the development of the FPSOs are concerned. Thus, in March 2015 Galp provided new guidance specific to the replicant FPSOs, where it stated that delays to the replicant FPSOs could be one year, on average, due to bottlenecks in the supply chain in Brazil. This guidance now looks sensible. Figure 18: Santos Basin FPSO unit schedule Santos Basin floating production, storage and offloading (FPSO) unit schedule FPSO Number Name: Cidade de Chartered/ owned Shell interest (%) Start-up[A] Capacity oil (kboe/d) Capacity gas (mmscf/d) Location 1 Angra dos Reis Lula Chartered onstream São Paulo Sapinhoá South Chartered onstream Paraty Lula North-east Chartered onstream IIhabela Sapinhoá North Chartered onstream Mangaratiba Iracema South Chartered onstream Itaguái Iracema North Chartered onstream Maricá Lula Alto Chartered onstream Saquarema Lula Central Chartered onstream Caraguatauba Lapa Chartered Replicant Lula South Owned Replicant Lula Ext. South Owned Replicant Lula North Owned Replicant Atapu South [B] Owned Replicant Berbigáo/Sururu [B] Owned Replicant Atapu North [B] Owned Lula West Libra Pilot Libra [A] Operator s view. [B] The Berbigão, Sururu and Atapu accumulations are subject to unitisation agreements. Source: RDS RDS in Brazil 9

10 Figure 19: PBR's current business plan to 2020 Source: PBR Since that guidance, however, mitigating measures, whereby more work is being transferred to Asian yards, are de-risking the outlook of the replicant FPSOs, potentially rendering Galp's guidance as slightly more conservative this is positive as further delays may now be less likely. These mitigating measures are possible as most licenses currently under development, especially those (ex-libra) to which RDS is exposed, have relatively low and manageable minimum local content requirements vs the local content applied to the replicant FPSOs by the operator, PBR, at the start of the development. For example, we believe the replicant FPSOs initially were sanctioned with a targeted local content of ~60%, when these licenses only are required to use ~30% from the local industry. Hence, mitigating measures can be implemented without significant penalties, in our view. Operation Car Wash is a complex investigation involving former PBR employees, executives of suppliers and politicians on alleged corruption charges within PBR, and has been running for more than 10 years. But similar to other partners of PBR, RDS likely views this as a short-to-medium term issue with longer-term benefits to PBR and the regulatory environment, which will likely include an enhancement of local content policies (there is a long process towards reliable offshore capability at local shipyards see chart on the following page), environmental licensing (often one of the key bottlenecks for oil and gas developments in Brazil), clearer fiscal and contractual rules (eg the Repetro regime), amongst other things. We discuss reforms in a later section of this report. PBR track record of late is looking better lessons learned from past challenges? The Santos Basin is strategic and the key driver of growth for PBR as the Campos Basin is mature and fast declining, so mitigating delays on these key FPSOs is important for PBR especially. As we stated earlier, it has done a good job bringing onstream the leased FPSOs in the pre-salt Santos basin over the past few years, but its track record is not perfect, in particular prior to the emergence of the pre-salt play in the Santos Basin. For illustrative purposes, we highlight a number of projects where PBR has not done as well as expected, which include: RDS in Brazil 10

11 The installation of the P-47 on Marlim (Campos Basin) was delayed by a year to end 2005 also due to issues with gaining environmental permits. Marlim Sul was delayed by more than two years due to late work on the conversion of production units and delays to environmental permitting. The first permanent production facility P-36 on the Roncador field (Campos Basin), which was in production since January 1999, encountered a natural gas exploration and sank in March Production was resumed in December 2002 via a leased FPSO. The Barracuda-Caratinga (Campos Basin) developments were delayed by more than a year due to problems with the work on the FPSOs related to contract specifications, contractor delivery, and local content issues. The Mexilhao (Santos Basin) development was also delayed several times due to delays in contracting the construction of the PMXL-1 platform and obtaining the required environmental approvals for the pipeline. PBR originally planned to start production at Mexilhao in July 2008, but first gas was only achieved in mid First oil from the Cascade/Chinook (US Gulf of Mexico) fields was also frequently pushed back. Work to complete commissioning of the FPSO has taken much longer than anticipated, which meant over a one year delay to the first oil target. Figure 20: Singapore yards a long process towards (reliable) offshore capability Source: Credit Suisse Research RDS in Brazil 11

12 Mitigating operational delays So far, PBR has done a good job mitigating major delays, and this was driven mostly by a more pragmatic approach towards local content (it has some flexibility around that on the licensed fields). In 2013, the consortium managed delays in the development of the replica FPSOs (locally built FPSOs) via contracting FPSOs from the international markets. This was not necessarily an acceleration of development for the licensed blocks (but instead to preserve the timeline) as the internationally leased FPSOs were swapped with the replica FPSOs with the incremental units deployed on the ToR fields (feasible due to the more homogenous characteristics of some of the fields). The ToR fields have more restrictive local content requirements 65% for the ToR fields versus 30% for BMS-9 and 11. Recent updates on mitigation delays includes the P-67 hull (Lula North), which has been transferred to China. The re-tendering of the IESA contract is another example. PBR also moved the contracts that the struggling Integra consortium had (OSX and Mendes Junior) to COOEC in China, which included the topside integration job for two of the replica FPSOs, amongst other things. Encouragingly, so far, these yards in China are doing a good job on these orders. Against this guidance, Galp on the 2Q16 conference call stated that five of the six units are largely de-risked, whilst the last still faces potential issues with the hull. There is plenty of time to risk manage this, in our view. Figure 21: Reducing execution risk Source: Galp (March 2016) Contracts, however, can't be simply cancelled. Upstream oil companies involved in the development of the Santos Basin are monitoring closely the execution of contracts (ie quarterly site visits). Typically, if a contractor underperforms, there are direct discussions to propose remediation/corrective action, which can include change in the work scope something difficult to sign off while companies remain under investigation (ie related to Operation Car Wash). Equally, the consortia cannot simply cancel contracts with contractors as 'under investigation' does not mean they are guilty, until proven. This is and can be an issue as companies rely on the performance of contractors and the financial strength to deliver on these, unless contractors are in breach of existing contracts. In other words, generally, in case of potential/imminent breach of contract or significant delays, it is fair to assume the consortium would try to negotiate and reach the best solution, where possible, in terms of project execution, which has been happening of late. Features of Long-life / Long plateau Data to date have shown higher well deliverability and greater reservoir connectivity which should allow for increased recovery per well. Results from EWTs (extended well tests) and the Lula Pilot have exceeded expectations with constant production, good behaviour of the RDS in Brazil 12

13 reservoirs, good lateral communication and no significant issues regarding flow guarantee to date. The planned FPSOs are positioned aerially over the discoveries and in view of better reservoir connectivity, continuity and deeper oil/water contact, the aerial coverage of FPSOs allows for a larger access to resources with the same number of wells (i.e., no incremental subsea/drilling capex). This allows for extended plateau production. BG previously stated that it expects to see better-than-expected connectivity elsewhere in the Basin, though admittedly this will vary from area to area as these are carbonate reservoirs. Figure 22: RDS pre-salt break-even analysis Figure 23: RDS entitlement production profile (kboed) $60 12, $50 10, $40 8, $30 6, $20 4, $10 2, $0 Sapinhoa Lula/Iracema Libra Greater Iara unitised Lapa E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E Break-even oil pirce (lhs) P50 rem. resources Lula/Iracema Sapinhoa Greater Iara area Lapa Libra BC-10 Source: Credit Suisse estimates Source: Credit Suisse estimates; Note: we assume additional FPSOs than currently guided to in the 2020s, most notably for the Greater Iara area. Figure 24: Lula Pilot extended plateau Figure 25: Lula Pilot: Galp now expects 7y at plateau 100% 90% ~18 months to reamp-up to plateau 80% 70% 60% 50% At plateau for ~4 years already 40% 30% 20% 10% 0% Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 Jul-13 Oct-13 Jan-14 Apr-14 Jul-14 Oct-14 Jan-15 Apr-15 Jul-15 Oct-15 Jan-16 Lula Pilot Source: ANP, Credit Suisse Research Source: Galp (March 2016) In mid-2011, BG stated that it expects to recover ~1bn boe from the location of the Lula Pilot. Initially, BG guided to recovery per facility of ~700mboe, which is also a significant number, and so far the signs are more than encouraging. This would indicate a much longer plateau and a lower decline rate. For example, Galp now thinks that Lula Pilot could produce at plateau for at least 7 years. Lula Pilot reached plateau in June 2012, and has been at plateau ever since, which translates to an annual FPSO utilization rate of ~90%. When we think of BG's Brazil assets as a whole (as opposed to FPSO to FPSO), there is the potential to sustain overall peak production at stable levels for a much longer period. RDS in Brazil 13

14 Improving recovery factor Big fields get bigger? Data to date have shown productivity exceeding expectations due to highly fractured reservoir coupled with greater reservoir connectivity, which will allow for increased recovery per well. Results from EWTs and the Lula Pilot have exceeded expectations with longer plateau duration increasingly likely. The potential on the Lula Pilot in terms of plateau duration has not been extrapolated, however, with Galp's base case (conservatively) for plateau duration at the other FPSOs being 2-3 years for now. These carbonate reservoirs have so far proven to be significantly better than expected. Carbonate reservoirs often provide excellent porosity, but are not always accompanied by great permeabilities; i.e. the pore spaces are not interconnected, restricting the ability of liquids to flow. This unpredictability in reservoir heterogeneity (negative effects on connectivity) is a key risk in petroleum production from carbonates. For example, Lula West and Lapa have turned out to be smaller in size after extensive appraisal, and Iara's reservoir varies in quality. That said, these barrels, bottom line, could potentially be even more long-lived than the bull case highlighted here if recovery factors move much higher. This is important as these represent low cost high value upside potential over time. Figure 26: Examples of improving recovery factor Figure 27: Deep Water Value Creation Source: BG Group (2011) Source: RDS November 2015 An upward moving trend. The recovery factor implicit in PBR's resource estimate of 8.3bn boe for Lula/Iracema (up from an initial 5-8bn boe) is 23%. The resource base is defined as 2P plus 2C, and in fact the recovery factor is measured on the Ps. Galp estimates Lula/Iracema to contain 9-11bn boe (2P+2C and 3P + 3C), and assumes a 28% recovery factor on 2P and its goal is to achieve ~40% longer term; something that PBR has often cited in the past as a possibility. On Sapinhoa, PBR estimates ~2.1bn boe of recoverable resources (up from an initial range of bn boe with the higher end at the time based on a 35% recovery factor) following the DoC (declaration of commerciality) in December Sapinhoa, with already high productivity potential of 50kbd per well, will not be a candidate for CO2 injection due to its already high permo-porosity and strong water drive (aquifer). On Iara, Galp stated at the time of the discovery in 2008 that the recovery factor assumed is ~16%, giving it a range of 3 4bn boe (gross) of recoverable resources. The initial calculation was based on a conventional solution for the development of the oil field using water injection. Following the declaration of commerciality in late 2014, the licensed part of Iara, however, is likely at the bottom end of the of 3-4bn boe, in our view. CO2 content Most of the fields in pre-salt Brazil have ~15 20% of associated gas, of which 10 20% is CO2 (Libra is seen to have ~45% CO2 content, while Jupiter is seen to have >70% CO2 content). In terms of field development, this associated gas can be used RDS in Brazil 14

15 for reinjection to increase the pressure in the reservoir, improving the recovery factor. During this process, if the oil and gas mix well (owing to the presence of CO2), the solution created helps to reduce oil viscosity as well as wettability. Advantages of lower viscosity oil is the increased ease to displace it from the reservoir rock. Thus, a combination of CO2 (improving oil viscosity) and water (helps the flow) injection improves the viscosity of the oil as well reducing the wettability, leading to an improvement in the recovery factor (3-17bpts, with the lower end already assumed by BG s guided P50 estimates). Issues that may be encountered is corrosion of risers due to the high CO2 content. In order to overcome this, alloys need to be developed for the risers that are able to withstand these conditions. Other issues can include the increasing amount of topside required for processing and compressing gas however, PBR's patented pressurized dense-phase separation technology (Hi Sep) could be a solution. The dense-phase extraction and handling processes being developed for Libra can allow CO2 injection at higher flow rates with potential to free up more of the topsides space for handling oil. Figure 28: Recovery factor upside on Lula/Iracema Figure 29: CO2 content Source: Galp (March 2016) Source: PBR Improving recovery factor from WAG WAG with CO2 floods was initiated on the Lula field in 2H12/2013 as part of the pilot. The purpose of a pilot is to gather data on the effectiveness of secondary recovery mechanisms, amongst other things. At the time, BG stated that it intends to apply WAG on a commercial scale from Ultimately the intention is to use WAG, as all the FPSOs have been designed to handle WAG. Historically, WAG can raise recovery factors by 3 17 percentage points, according to BG. The first test/process started in September 2012 (with water injection), according to BG. The process involves the injection of water into the WAG-CO2 system for six months or so before switching to CO2 (in late 2Q13, according to BG) for another six months of testing. In the future Galp expects this to be periods of three months each. Whilst companies stated in late 2013 that results had been encouraging with the first well producing dry oil, without gas or water breakthrough or any significant pressure loss, definite answers are likely to take more time for various reasons. There is no update yet from the companies. Generally, it will be a step-by-step approach; to see if it is operationally effective (in essence, a CO2 based WAG injection had never been carried out on deep-buried carbonate reservoirs), and only after one year, the consortium should see the first effects on the producing well, which means, it could take a number of years until we can have a definite answer (ie whether the recovery factor can indeed improve). RDS in Brazil 15

16 Figure 30: CO2 injection in the reservoir Figure 31: Lula horizontal well Source: Galp (2011) Source: PBR Horizontal drilling The first horizontal well on Lula was successfully concluded in March 2013, but not much has been disclosed so far. What we know is that this well was a test, which was successfully concluded, and is producing in that area. We remind investors that the Lula Pilot is a 'pilot' area, where the consortium tests solutions like drilling, operation and potential benefits of having deviated wells in this carbonate reservoir. Galp stated that, although this well was successful, in the current 'initial' development stage of the Lula/Iracema field, the consortium is less likely to add additional horizontal wells. In other words, the current drilling plan is based on vertical wells, which have shown extremely positive flow rates, but horizontal wells should not be entirely disregarded, in particular in areas of the pre-salt, where improvement in productivity is desired and/or necessary (eg parts of the Greater Iara field, in our view). Figure 32: Extracting full value Source: Galp (March 2016) RDS in Brazil 16

17 The deviation of wells through salt is also a significant technical challenge. The challenges of drilling through a thick salt layer are caused through the movability of salt, and the difficulty in controlling the integrity of the well. By its nature, salt can move and deform which presents a number of geo-mechanical issues leading to instability and casing deformation. Key variables that contribute to successful directional drilling in salt include hole size and geometry, wellbore trajectory, casing design and effective cementing. The industry has found techniques to temper the challenges, i.e. increasing the ROP (rates of penetration) using rotary steerable systems, real-time monitoring, and decision making. Acid injection can significantly boost production in carbonate oil and gas fields as the reaction between an acid and calcite or dolomite can enhance flow properties of the rock. To hydraulically fracture carbonate rocks, acid is pumped into the fracture to create an imprinted pattern on the surface. The acid then creates conductive wormholes. Ramp up period Continuous improvement in the ramp-up period. The base case for the consortium is to ramp-up to plateau over a month period, but some of PBR's partners (eg Galp) have an ambition to deliver this in just 12 months. For example, we have seen continuous improvement in the delivery on the Lula/Iracema field. The first leased FPSO took 18 months to reach plateau from first oil, the second took 15 months and the third took only 13 months. The fourth FPSO (150kbd capacity), which started up in July 2015, would have taken only 12 months to reach plateau production, if it wasn t for the small delay with the connection to the gas export network. Generally speaking, with the Cabiunas gas pipeline now operational (since February 2016), albeit with some operational hiccups since, lack of gas evacuation should not be an issue for now. The key reasons for a faster ramp-up include: (a) better flow rates (i.e. fewer wells needed), (b) good well inventory (pre-drilled wells), and (c) faster hook-up (e.g. availability of PLSVs and learning curve). In more detail, the variables to help achieve this potential medium to longer term target are as follows: Rig availability: more rigs are now available, which means it can pre-drill more wells. Drilling efficiency: Across Lula/Iracema, the average drilling and completion time in 2015 stood at 110 days with the best well taking just 75 days (which is Galp's longerterm ambition, but not its base case, which is likely closer to ~105 days and consistent with our estimate of ~100 days for the Lula/Iracema field). Galp stated on the 2Q16 call that 2016 to date has seen another ~20% improvement versus 2015, which is impressive, and PBR at a recent industry event in Rio in late August 2016 said it has drilled and completed three wells in the pre-salt in ~50 days each. We have previously argued that investors should not extrapolate the best wells drilled because some parts of the field (eg, the Iracema reservoir) are easier to drill and the best ones may have also had no weather-related outages (and may have also been merely an open hole, single casing, water injection well). That said, the learning curve over the past few years have been impressive as far as the drilling on Lula and Sapinhoa is concerned. Well count: productivity of the wells in the Santos basin are better than expected 25kbd in the Santos Basin vs ~15kbd in the North Sea vs ~10kbd in the US GoM. This means fewer wells need to be hooked-up to reach plateau. Overall, the number of wells to be drilled is likely to be unchanged, in our view, but the well productivity and the reservoir connectivity means that the recovery per well will be higher, and thereby extending the plateau duration of the production unit. Pipelay support vessel (PLSV) availability: this has been a bottleneck for some time, and PBR has been more actively ordering PLSVs. More pipelay support vessels means that it could reduce the hook-up time. RDS in Brazil 17

18 Figure 33: Drilling and Completion time evolution in the Brazilian pre-salt Source: PBR Figure 34: Galp specific: Drilling and Completion times Figure 35: Lula/Iracema outstanding productivity Source: Galp (March 2016) Source: Galp (March 2016) Well performance Campos (15kbd) vs Santos (25kbd) Pre-salt well flow rates in the Campos Basin are averaging ~15kbd. This is a good number but no match to the average ~25kbd in the Santos Basin this compares with ~15kbd in the North Sea and 10kbd in the US GoM. Some of the better wells flow (excluding gas) at 40kbd plus. The pre-salt reservoirs are so far showing exceptional hydraulic communication dismissing the concern of a steep production decline curve because of the compartmentalised nature of most carbonate reservoirs globally. The worst performing well is the second producer on Lula Pilot. Structurally, this well will not produce more than 15kbd of oil. This is because of the reservoir characteristics at the well location, which is not as productive as the others. Yes, in carbonate reservoirs, even short distances can have an impact on the characteristics (ie not completely homogenous). Either way, Lula Pilot managed to reach plateau from only 4 wells and these wells can sustain plateau, but the consortium is currently producing from 6 wells. The flow rates are being chocked as a result due to the limited size of the production unit. RDS in Brazil 18

19 Well performances are broadly speaking exceeding expectations of the operators (in line with our estimates), which would lower capex/boe relevant as 50% of capex is for drilling and completion. That said, the consortium will still drill the same number of wells as planned. The well performance, however, will allow for a longer plateau duration with the same number of wells, which is consistent with our estimates. Drilling time Progress has been made in reducing drilling time and costs. The average time to build a well reached 89 days, a reduction of 71% between 2010 and 2016, according to PBR. The success of drilling the Brazilian pre-salt reservoirs requires a broad range of technologies. The reservoirs are complex variable layers of carbonates and distinguishing reservoir characteristics is difficult. In carbonate formations, drill bits undergo significant impact and heat damage as well as abrasion. The bottom 20% of the well can take up to 80% of the drilling time with low ROP. PBR has come across these problems in Brazil, having to change drill bits 2-3 times every 200m, which increased rig time and associated cost. PBR, the operator of the Santos Basin, is working with contractors to research different technologies to increase the ROP in the hard pre-salt rock. Wells that were initially drilled in five or six phases are being drilled in fewer phases (three or four), yet with the same final diameter. The RoP in the pre-salt carbonate rocks have gone from one metre per hour to as much as three metres per hour, according to PBR in Broadly speaking, the improvement in the drill time has been mostly driven by learning curve and improved well design as opposed to through the introduction of new technology. Potential introduction of ground-breaking technologies can help further, which may benefit the later field developments, such as Libra, and the Greater Iara area. Gas monetization At the start of 2011, BG stated that total gas resources in the Santos Basin were more than 14tcf. Today this figure should be much higher. PBR at the time was considering three to four gas offloading routes. The first route is via the completed Lula Mexilhao Caraguatatuba pipeline with capacity of up to 10m m3/day of gas (originally for 3 FPSOs); the second route is via pipeline through Cabiunas (for 13m m3/day), where the gas processing capability is already being upgraded (originally for 3 4 FPSOs); a third route using another gas pipeline dubbed Route 3 (c18m m3/d capacity or originally for 5 6 FPSOs); and a fourth route similar to the third but with 14m m3/day. One of the reasons why the Santos Basin is strategic to Brazil (aside from the very large recoverable oil resources) is also that it has big gas resources (even adjusted for reinjections) associated to these oil fields, and Brazil wants to monetise these in the domestic market. This means that gas infrastructure is needed. Evolving gas evacuation plan: under the normal or initial plan, the Cabiunas pipeline (or Route 2) needed to be in place by mid-2014 to accommodate for FPSOs 4 and 5 (but it became only operational in February 2016). This is because the Santos Basin is generally more gassy (associated gas production of ~20%) and given restrictions to flaring, the absence of an export route would in theory impact the oil production curve. Despite the delay to the completion of the Cabiunas pipeline, PBR stated the impact to oil production would be limited because, even with a one year delay to the completion, it has room to increase re-injection within the technical limits. This meant there was more spare capacity in the Mexilhao line than the base case, and production indeed had not been impacted. Developing subsea gas pipelines offshore Brazil has not been a straight forward process. The delays to the Cabiunas pipeline were multi-factored, which included things such as missing environment permits from Ibama, delays to the supply of the pipes. At this stage, the Rota 3 has been delayed from the initial plan (now targeted start-up is 2019). Partly, this is due to the environment delays with Rota 2 PBR reviewed the shallow-water part of Rota 3 with the aim to simplify the project to mitigate environment risks. It was generally RDS in Brazil 19

20 looking to simplify the project as prices were also seen as being too high. The company's Engineering department concluded the technical simplification studies, and thereby reduced pipe diameter from 24'' to 20'' in 197km between 1,500 and 2,300m water depth, and also reduced the number of PLETs to five (from seven previously). Figure 36: Santos Basin and gas pipelines Source: BG Group Unitisation challenges There are a number of areas in the pre-salt Santos basin, which are subject to unitisation. Potentially more will emerge and this is due to the favourable geological characteristics of the pre-salt in the Santos Basin with large, numerous and interconnected reservoirs. These processes come with challenges, more so than in other countries, and one of the more challenging one looks to be the Greater Iara field, in our view. Generally, even if the operator is the same, the process is never a straight forward exercise, especially if the fiscal terms, local content requirements, license period etc are different between the unitisable areas. The TOR fields also have different fiscal terms as PBR at the time of the acquisition pre-paid for the SPT (Special Participation Tax); in other words, once these TOR fields come into production, they are not subject to the SPT. The unlicensed areas will be based on Production Sharing Contracts (PSC); similar to Libra, based on the current regulation in Brazil. RDS in Brazil 20

21 Rule of Capture versus Unitisation In the earlier days of the oil industry, in the US, the concept of unitisation did not exist; a concept called Rule of Capture prevailed instead. The Rule of Capture stated that a land owner, if able to produce oil in his property, would be entitled to do it and to have the rights to that oil, regardless of whether the oil reservoir from which it was producing extended or not to a property nearby. The industry soon discovered, however, that the rule of capture did not yield the best economic or geological results. It incentivised the over-exploitation of the reservoir by the various land owners, and therefore early reservoir depletion. The Spindletop field is a classic case-study: it was discovered in 1901 in Texas, with the first well producing an average of 100kbd. By 1902, there were 285 wells in operation, and in 1903 the uncontrolled production led to the first decline rates observed in the field. By 1910, the field was already depleted. The concept of unitisation was soon developed to allow for adequate exploitation of the oil fields, maximisation of the recovery factor, and cost advantages in sharing infrastructure. Brazilian historical context Since 1997, when PBR s monopoly for oil and gas exploration and production in Brazil was terminated, four processes of unitization have been initiated and completed. We discuss below the main aspects of each of those process. The discussion is based on a technical paper by Bonolo and de Almeida, petroleum engineers at PBR. Lorena x Pardal. Lorena is a field that had been producing since 1985, but had its production interrupted in 2002 after it was identified that the accumulation extrapolated the limits of the concession. At the time, the nearby area was being bid in the context of the fourth licensing round, ultimately acquired by Potioleo. Negotiations went on for seven years, and in the meantime, Potioleo declared the commerciality of the Pardal field. In 2009 the companies submitted the Unit Agreement, which resulted with 73.9% of the accumulation in Lorena and 26.1% in Pardal. One of the key points of discussion was how to treat different local content (LC) rules: Lorena did not have minimum LC requirements, whereas Pardal had a minimum of 70% LC in the development phase. An agreement was reached to weight up the local content by the respective volumes, with a resulting 18.27% for the unitised accumulation. Albacora x Albacora Leste. Albacora Leste had been producing since A development well drilled in 2006 found out that one of its accumulations (called Arenito Caratinga) extended to the nearby concession Albacora. The Unit Agreement was submitted in May 2007, having been approved in December the same year. There are a number of reasons for the significant different pace of unitisation in this case (less than a year) vs the Lorena-Pardal (7 years), such as the same operator in Albacora s case, with extensive geological knowledge of the accumulation, and prior experience obtained by the operator and the ANP with the first unitisation. We also note that the only accumulation unitised was Arenito Caratinga, responsible for 8% of the total production in Albacora and Albacora Leste. Mangaga x Nautilus. This was the first time when unitisation occurred within two commercial areas with different operators (RDS in Nautilus and PBR in Mangaga). The unitisation involved three different accumulations in both fields, and the final split was 50/50. Interestingly, PBR was the company with the highest stake in the unitised accumulation, but RDS was chosen as the operator as it already had nearby infrastructure available at the Parque das Conchas complex. Camarupin x Camarupin Norte. Similarly to the Albacora and Albacora Leste cases, the unitisation of Camarupin and Camarupin Norte was made easier by the fact that PBR was the single operator in both areas. Also, like for Lorena and Pardal, there were different local content requirements: 0% in Camarupin, and 30% in Camarupin Norte. The unitised local content was also a weighted volumetric average: 20.85% LC equivalent to 30% LC in Camarupin Norte times 69.5% volumetric weight. RDS in Brazil 21

22 Pre-salt unitisation: potential challenges Since the end of 2010, the Brazilian pre-salt province has been subject to three different regulatory regimes: concessions, transfer-of-rights, and PSCs. And because of the favourable geological characteristics of the pre-salt, with large, numerous and interconnected reservoirs, a number of industry specialists believe in a high-likelihood that unitisation in the area is set to increase. Unitisation on its own is not necessarily a simple issue to get around, with relevant processes globally taking four to nine years to get resolved. The key implication of this is a risk of delayed development. In the pre-salt case, the complication is the presence of three different contractual structures, most of which (~70%, PSC areas) are still unlicensed. Below we discuss a number of matters which we think are the most relevant points when it comes to pre-salt unitisation: Figure 37: Iara (licensed) and Entorno de Iara areas (Transfer of Rights) Source: Galp Definition of specific contractual rules in areas containing different contractual structures. This issue arises from Article 36 2, Law /2010, which mentions The exploration and production regime to be adopted (in unitized areas extending to unlicensed acreage) depends from the regime adopted in adjacent areas. 'Divisible obligations are those that can be met separately by different parties involved. For instance, if fiscal terms differ across different areas to be unitized, they could be applied separately to each oil company in the different areas, maintaining each company s original contract. This is not in the law yet (the law, as per above, mentions a specific contractual structure will be created for the unitized area). However, there is one official interpretation that mentions the fiscal terms in unitized areas must be applied independently and proportionally in a way to respect the characteristics of the original contracts signed by the companies. Indivisible obligations are those that cannot be met separately by the different parties involved. The most notable example here is local content: even if there are many local content requirements in the various areas to be unitized, only one final LC can be achieved. So far the most common solution has been towards a proportionate allocation. For example, the LC for BMS-11 is 30% versus 65% for the TOR fields. RDS in Brazil 22

23 Operations continuity. Article 41, Law /2010 states that The development and production of the accumulation will be suspended until the Unit Agreement is approved by the ANP, unless in specific conditions authorized by the ANP. The Unit Agreement is submitted by the oil companies involved in the unitization, and defines the percentage stakes of each company in the unitized accumulation, the operator, the development plan, redetermination periods, etc. The role of PPSA (Pre-Sal Petroleo S.A.) in the process. PPSA was created by Law /2010, August 2010, to represent the Federal Government on a number of processes related to the management of PSCs and commercialisation of oil and gas. PPSA will also act on behalf of the government to set the Unit Agreement with interested parties if accumulations spill over to unlicensed acreage in the pre-salt area. Once an agreement is struck, it still needs to be approved by the ANP, the Energy Regulator. The president of PPSA acknowledged that working out these agreements will be one of the biggest challenges PPSA will face, particularly as different fiscal terms live side by side. Failure to reach an agreement in a timely manner may have the potential to slow down development plans, in our view. License period The question clearly is whether oil companies will be able to produce the total resource base within the license period. That depends on the resource estimates, and using BG's estimates this looks less likely, and will depend on how fast further FPSOs are deployed and how much each FPSO can recover (ie more than the base case?). As part of the contract, the consortium has the right for extension subject to approval by the ANP after the 27-year production period expires. It is common that the existing/operators players receive an extension, but it can also come in a slightly different form. A number of contracts in Asia, when renewed, were awarded entirely to the national oil company with the international oil companies left out. Or it can also come in the form of less attractive fiscal terms upon renewal/license extensions, and lower equity stakes given to existing contractors, and no guarantee, as other players may be able to compete against existing players. That said, interestingly, recently upstreamonline (5 February 2016) highlighted that the government is drafting a proposal designed to extend licenses with greater flexibility on local content and prolong the Repetro tax exemption for another 20 years. As we understand it, the areas identified to prioritise the extension discussion includes the Campos basin and a few others (but not the Santos basin) as these are the more mature fields that need extension to secure the investments needed to sustain production. In other words, it is still too early to discuss extensions on the Santos Basin. What about the Rio taxes? At the start of the year, the state of Rio de Janeiro independently from the Federal Government implemented two additional taxes, which include (a) a fixed tax, which would amount to $0.69/boe on production in the state (much of the Santos Basin, excluding BM-S-9, is in the state of Rio de Janeiro) not too big a burden and (b) incremental 18% royalty on sales, which is significant. The IBP (industry association) and companies independently are opposing the taxes introduced by the State of Rio our base case is that the ICMS will be suspended (similar to 2003) before April 2016 as it is unconstitutional. An identical bill to ICMS (VAT) was proposed in Rio de Janeiro State (the so called Noel s law ) in A lawsuit arguing the unconstitutionality of Noel s Law was filed and a decree was enacted suspending the effects of such law. Such suspension was derived from the solid legal arguments of the industry to invalidate such law, given constitutionality and legality issues. Therefore, no ICMS has been charged in respect to crude oil extraction. Galp believes that the Supreme Court will rule in the next 6-12 months; until such a decision, there is a payment injunction (ie no payment will be made to the state of Rio related to these taxes). Galp also stated that so far it is getting positive signals. Our base RDS in Brazil 23

24 case is that the ICMS will be considered unconstitutional. Moreover, the application of such a proposal is detrimental to the oil industry in Brazil and especially to the NOC, PBR. Libra The ANP estimates recoverable resources of 8-12bn boe with the lower end of the range representing Gaffney Cline's best estimate from 2010 (3.3-15bn boe on a % recovery factor), which at the time stated relatively minor concerns over trap and seal effectiveness, plus reservoir presence and quality. For Total and RDS, having taken a 20% stake each in the field for a price tag of $1.4bn would imply a finding cost of $ /boe not a bad price. Only one well had been drilled at the time of the bid round, and perhaps the lack of data on Libra and quality of seismic data in theory would make such an estimate on the resource base hard to believe. These are carbonate reservoirs (ie, can be highly heterogenous), but we think that the ANP would have done the correct extrapolation to derive its figure. We note that 60 appraisal wells, 41 DSTs, 1.5km of core samples, 72 conventional logs and 57 production logs had been done in the pre-salt between 2006 and the time of the bid round on Libra. Big numbers that are crucial for PBR to keep enhancing its understanding of the pre-salt. The DoC deadline is a maximum of four years after contract signatures (or late 2017) with first oil required to be delivered within a five-year period following the DoC. The first large scale FPSO is targeted to be brought onstream in 2020 with many more thereafter, and the pace and shape will depend on PBR's next Business Plan (September 2016). At the time of the bid round, the ANP mentioned a peak gross production potential of 1.4mbd. Ever since Libra was auctioned and awarded in October 2013, many signals continue to indicate that this is a project that is being well managed and going well: Since December 2013 two months after the award PBR created a special 'Gerencia Executiva' (a separate organizational unit reporting directly to the Director of E&P) for Libra. The operational committee of Libra approved on time (end January) the 2014 investment plan for the block, and drilling activity started on time. The consortium is also well ahead to securing the FPSO for the extended well test in Libra, which should take place in The FPSO will have a capacity of 50kbd of oil and 4MM3/d of gas. Figure 38: Libra base case when announced in 2013 Source: Credit Suisse Research RDS in Brazil 24

25 Access to more resources When the oil price is low, it is tough to be thinking about spending money that you don't have, but it is times like these when good opportunities emerge. Slowly but surely more distressed companies realise that only part of their crown jewels will attract the cash constrained industry to consider spending the precious limited capital available. Libra (pre- FID) is precious and we have previously argued PBR could sell 10% (which takes it to 30% or the minimum level as per regulations for this license). Recently, the Brazilian Senate passed a bill that exempts PBR from the minimum mandatory stake of 30% and operatorship in all pre-salt fields (the bill has to go through the lower house yet), which could mean that perhaps more than 10% in Libra could be monetized. It would be natural for existing partners of PBR to look into potentially increasing equity in this license, which include RDS, Total, CNOOC and CNPC. We have recently seen PBR sell its 66% stake in the Carcara discovery, offshore Santos Basin, to STL. Other opportunities will likely emerge over time. Exploration in Brazil The last 'relevant' license round in Brazil is now a number of years ago, and much of the focus then was on the Equatorial Margin. Operating in the awarded Equatorial Margin blocks will be challenging, according to PBR. Strong currents, environmental issues and lack of infrastructure in the region will all need solving. When asked about those challenges, the former head of PBR E&P, Mr Formigli, reminded the audience that PBR will not operate any of the blocks in the Foz do Amazonas and Barreirinhas basins, and that it will be a 'privilege' to see how operators BG and Total will address those issues. According to BG (now owned by RDS), it opted for the Barreirinhas basin as it believes that the external operating environment is less challenging than the Foz do Amazonas (the Foz do Amazonas were more highly bid given the proximity to French Guiana). These blocks have large Cretaceous fan plays, analogous to Ghana, that are fed by Cenomanian/Turonian source rocks. As well as being trapped stratigraphically, BG believes that these features also have roll-overs updip that reduces the risk of leakage associated with stratigraphic trapping mechanisms and provides much better integrity. Reforms Around 10 years after the discovery of the pre-salt, one of the most relevant events in the global oil industry since 2000s, the following has occurred: Brazil and PBR have still not meaningfully grown production, with a direct impact to the companies involved, the country/states/municipalities which depend on oil-revenue for their budgets, and even other sectors unrelated to oil (a portion of the pre-salt royalties have to be invested in the Education sector in Brazil). PBR, the company responsible for spearheading the development of the new resources, is in a much worse financial situation than prior to the discovery of those resources. The local industry has yet to climb the learning curve to be able to be competitive at an international level, and now many local suppliers find themselves filing for bankruptcy or judicial-assisted-recovery. Few other oil companies have meaningfully increased activity in Brazil. The lack of regular oil licensing rounds and the regulation requirement for PBR to be the sole operator in new pre-salt areas has impaired the rise of other oil companies in Brazil. Addressing the above shortcomes means the Brazilian Oil model needs reforms on a number of fronts. The Brazilian Petroleum Institute (IBP), a non-profit private organization founded in 1957 that today comprises over 200 associated companies, is one of the most active and representative bodies of the oil & gas industry in Brazil. The IBP has been RDS in Brazil 25

26 pushing for reforms in the oil sector on a number of fronts that in our view would have a long-term benefit for the sector, the country, and all the companies involved. Some of the key reforms being pushed by the IBP include: Frequent licensing rounds: adoption of a long-term calendar of licensing rounds by the ANP (Brazilian Oil regulator). Visibility and predictability allows for long-term establishment of operations of oil companies in Brazil, boosting investments, employment, and government revenues. The lack of frequent licensing rounds has been a key reason for the sector being overly dependent on PBR, and for the lack of a sizeable junior E&P industry in Brazil. Reviewing the PSC model to allow for (a) multiple operators, removing the sole-pbroperatorship-requirement, (b) better relationship between the veto and voting power of the PPSA (the government-owned company responsible for monitoring the budget and operations of the consortia), (c) improve the cost-recovery mechanism for the oil companies. Enhancement of local content policies. When the current local content policy was established, its initial challenge was to ensure adequate utilisation and development of the local suppliers in the oil and gas industry. Currently, local suppliers have a strong backlog but continue to face execution issues to the point where many of them face financial difficulties. The local supplier is not becoming competitive vs the international markets, and all stakeholders are seeing the consequences of delays. R&D. Current E&P contracts establish that 1% of gross revenues from oil and gas production subject to special-participation tax be invested in R&D. So far, those resources have been mostly invested in research centers in universities. This has enabled the creation of over 800 oil & gas research labs in Brazil, but there still remains a suboptimal active participation of the oil companies and suppliers in the process. Allowing more R&D investments targeted to the supply chain could boost the efficiency of such investments. Environmental licensing is still one of the key bottlenecks for oil & gas development in Brazil. This is increasingly important now that companies are starting exploration efforts in the equatorial margin North Brazil, an area with strong currents and complex environmental licensing requirements. Over the past two years, there has been a small euphoria regarding the unconventional potential of Brazil, especially in the Sao Francisco Basin. Outside the geological potential itself, the lack of adequate infrastructure and clear environmental licensing rules has been a bottleneck for the beginning of the unconventional efforts in Brazil. Clear fiscal and contractual rules are crucial to induce investment in the oil & gas segment, a long-term, capital intensive industry. Some challenges faced currently by oil and services companies in Brazil that need improvement are: (1) the Repetro regime a special import-export regime that lowers import tax for oil and gas equipment is due to expire by 2020; (2) change in interpretation from the tax treatment of offshore services leasing and bareboat charters; (3) involvement of the Brazilian anti-trust authority in the farm-in and farm-outs operations, beyond what had previously been required by the ANP, (4) change in the rules of calculation of the Special Participation Tax ANP resolution 12/2014 removed some expense items that were previously deductible for SPT calculation purposes therefore changing the economics of already approved and ongoing E&P developments; (5) changing (or threatening to change and increase) taxation by states and municipalities in the oil and gas production activities. Natural gas market. (1) Enhancing dialogues between the oil and gas and power generation sectors to ensure more constant utilisation of gas-fired power plants, giving visibility and stability to natural gas producers, and reducing LNG imports in Brazil; (2) Better integrating E&P licensing rounds, energy auctions, and investments in gas RDS in Brazil 26

27 pipelines as planned by the PEMAT government plan to increasing the gas pipeline network in Brazil. Bottom line, two examples which we think are most relevant for investors to understand: (a) the review of the PSC model, notably the removal of the mandatory-pbr operatorship, has a very limited impact on Petrobras cash flows, earnings and capex requirements before 2020, and (b) the enhancement of local content policies by, for instance, allowing for more flexibility and use of foreign-suppliers, can reduce the risks of delays and downgrades of PBR's oil production curve, can reduce PBR capex needs, and could be a positive for international shipyards in terms of new orders. Reforms should be forthcoming. The Brazilian government is sending positive signals as far as reforms are concerned. Whilst early days, the government is drafting a proposal designed to extend licenses with greater flexibility on local content. The simplification of the unitisation process involving different fiscal terms is another aim, and the argument for simply stretching the 'concession' (tax and royalty) regime to cover reservoir extension is compelling in cases where only a small portion of the reservoir extends outside of the license. Recently, the Brazilian Senate passed a bill that exempts PBR from the minimum mandatory stake of 30% and operatorship in all pre-salt fields. Alternatively, according to the amended bill, as it was approved, PBR will have the right of preference to the 30% stake and operatorship. The bill will now go to the lower house. The end of the mandatory stake and operatorship in all pre-salt fields, if it becomes Law, is positive for PBR shareholders in the medium and long term, in our view, because it reduces the risk of future mandatory capital commitments. Furthermore, the right of preference to a 30% stake and operatorship could be exercised by the company if there is an attractive pre-salt round and if it finds itself with a healthier balance sheet in the future. RDS in Brazil 27

28 Key Upstream Asset description Figure 39: Pre-salt in Brazil overview of key blocks and companies involved Source: Credit Suisse Research based on PBR, ANP; Note: STL announced the acquisition of PBR's stake in Carcara, which is due to close during 3Q16 Lula/Iracema (BM-S-11) The pre-salt area around Lula is known as the Polo Pre-salt (or Cluster Area) and is host to many major discoveries including Iara and Sapinhoa. Commerciality was declared on Tupi on 29 December 2010, at which point the field was renamed Lula. Lula delivered first oil within four years (in late 2010) after the initial discovery an impressive track record. BMS-11 has delivered three large-scale discoveries, namely Lula, Iracema and Iara. Lula (formerly Tupi) and Iracema were declared commercial on 29 December 2010, upon which PBR revised its resource estimate to 8.3bn boe (from a range of 5-8bn boe previously for Tupi/Iracema). At the time of the declaration of commerciality, PBR estimated recoverable resources of 8.3bn boe. Subsequently, BG raised its estimates further (not disclosed per field) due to higher/incremental in-place volumes on the flanks and slightly higher recovery factor (now assumes the low end of the 3 17bps upside potential to recovery factor from WAG). Overall, drillstem tests (DST) from the pre-salt Cluster Area has been very promising to date with IP rates of 30 50kbd considered possible. Iracema showed initial flow rates of up to 50kbd with low CO2 content. This compares with initial expectations of an average initial well flow of 15 20kbd and now moved up to 25kbd-plus for many of the fields in the pre-salt Santos Basin. Lula also had small set-backs. For example, Galp revised downwards the reserves on the Lula field in early The appraisal campaign on Lula West showed lower porosity, thereby reducing the in place volumes impacting the recoverable reserves by ~600mmboe (gross). Galp now estimates Lula/Iracema to contain 9-11bn boe (2P+2C and 3P + 3C), and assumes a 28% recovery factor on 2P and its goal is to achieve ~40% longer term. Lula vs Iracema. The declaration of commerciality (DoC) for Lula/Iracema in late 2010 also revealed that Lula and Iracema are separate accumulations, according to the consortium. Previously it was thought that Iracema was a northwesterly extension of the main Lula field. As a result, discrete blocks had been drawn around the two fields to be treated as separate tax ring fences. Upon the declaration of commerciality, PBR submitted a development plan to the ANP for consideration. RDS in Brazil 28

29 The ANP has officially declined to allow Lula and Iracema to be considered two fields. It was clear in stating that when the ring fence of the area was made, both structural highs related to Lula and Iracema were considered part of the same structure and therefore it believes it is not correct to change this understanding upon the DoC. The discussion is relevant because, when treated as separate fields, the consortium will end up paying a lower government take (tax) than if both fields were considered part of the same field (the special participation tax is variable to production levels and can thus vary depending on the pace of development and the size of the field). According to ANP at the time (in 2014), the tax difference may amount to R$50bn. Greater Iara Area (BM-S-11) The Iara discovery was announced in August 2008, some 60km northeast of the Tupi discovery and 230km off the coast. PBR announced recoverable resources of 3 4bn boe, and Galp stated that this is based on a 16% recovery factor (implying that the oil-in-place is similar to the giant Lula field). After re-entering the discovery well in the second half of 2009 PBR confirmed the initial resource estimates; however, the well could not complete the formation test. What may have made PBR more positive on Iara is when an appraisal well (Iara Horst) was drilled. The initial results following the successful appraisal well earlier in 2012 demonstrated superior reservoir characteristics compared with the original discovery well located ~8km away, but subsequently the flow tests disappointed. A number of flow tests on the central part of the field showed 5-10kbd unconstrained rates, and thereby questioning the economics of certain locations of the Iara field. As such, Iara's development was overtaken by Sapinhoa and the ANP granted an extension to the DoC deadline as more work was needed on the Iara field. Since then, drilling of the western flank has proven more promising. The western flank of this large reservoir had proven to have similar reservoir and fluid properties as Lula with the Extended Wel Test (EWT) running at ~29kboed (oil and gas). While the lower gas content on Iara could lead to lower pressure in the reservoir, PBR seemed confident enough at the time to stick with its 3-4bn boe volume guidance following the appraisal wells on Iara Horst and Iara West. While PBR's consortium partners turned more positive on the recoverable resource potential, Iara has proven to be much more heterogeneous than other fields in the Santos Basin; generally a common risk of carbonate reservoirs. Figure 40: Iara (licensed) and Entorno de Iara areas (Transfer of Rights) Source: Galp RDS in Brazil 29

30 At the time of the DoC, the operator PBR released a revised resource estimate for the Iara (licensed) and Entorno de Iara (ToR and SVToR) areas, and stated that the wider Iara area to contain 'over 5bn boe'. It has not provided an exact split, but this looked a somewhat less upbeat update than before. The ANP's update on Entorno de Iara in mid showed the area (ex the concession area) should contain a minimum of 3.1bn boe, which we calculate as 0.6bn boe as part of the ToR terms and bn boe for the Surplus Volumes for the ToR terms (SVToR). As this is the latest update specific for part of the wider Iara field, the licensed area would then amount to less than 3bn boe. This assumes the wider Iara area containing 5-6bn boe recoverable resources. In other words, the licensed Iara area would fall short of PBR's prior guidance of 3-4bn boe. That said, Galp said that it still believed (based on its own estimate) that the licensed area falls within the old range and is comfortable at the low end of the range of 3bn boe. Complex structure. All three newly defined areas are different in reservoir quality, and thus it is difficult to gauge the exact split by area. Despite having submitted the DoC for the wider Iara field as required by the deadline provided for the submission of the DoC, we think still a lot of work is needed for the three accumulations, most notably the central part of the wider Iara field, given the heterogeneity. Indeed, PBR stated in its recent release that 'subject to a complementary campaign for data acquisition, additional production units could be contracted by the consortium'; the success of which we assume. Ring fencing: consortium pushing for three separate ring fenced areas With the submission of the DoC, the consortium has also proposed the creation of three separate ring fenced areas for the wider Iara field. This is relevant as it would increase contractor share (or lower government tax-take) due to a lower Special Participation Tax (SPT). The SPT is a tax charged progressively to production levels. It is a proposal and thus not given that the ANP approves it as such proposals (eg Lula and Iracema) in the past have been rejected. Local content differences could complicate the unitisation process There are differences in local content for the licensed, ToR and SVToR areas. The license area has a minimum local content requirement of 30% during the development period, while the ToR is closer to 60%. The new SVToR areas will likely have similar requirements to Libra, which would amount to a ~60% local content requirement. With each FPSO being 'shared' in the wider area, it will be interesting to see what arrangement will be reached for the field development. o This was one of the key points of discussion for the Lorena and Pardal accumulation in the past, which had different local content rules. Lorena did not have minimum local content requirements, whereas Pardal had a minimum 70% local content in the development phase. An agreement was reached to weigh up the local content by the respective volumes with a resulting 18.27% for the unitised accumulation. Fiscal terms differ across the license, ToR and SVToR areas Without factoring in the upfront payments PBR had to make for the acquisition of certain resources, ToR terms are the most attractive as no SPT is incurred, followed by the licensed concession area, then followed by the SVToR area. The SVToR will follow the PSC model with PBR claiming that the development of these incremental barrels will yield returns similar to those of Libra. o Divisible obligations are those that can be met separately by different parties involved. For instance, if fiscal terms differ across different areas to be unitised, they could be applied separately to each oil company in the different areas, maintaining each company s original contract. This is not in the law yet (the law, as per above, mentions a specific contractual structure will be created for the unitised area). However, there is one RDS in Brazil 30

31 official interpretation that mentions the fiscal terms in unitised areas must be applied independently and proportionally in a way to respect the characteristics of the original contracts signed by the companies. License period is different for the licensed, ToR and SVToR areas Once a field is declared commercial, the production period of 27 years begins for a licensed concession area. The ToR area's production period lasts for 40 years from September 2010, but based on the example below, the end of the production of contracted volume stops in This will clearly vary from area to area and in the case of Entorno de Iara, it is likely earlier as the ToR volumes for the field amounts to only 0.6bn boe (the SVToR portion amounts to bn boe). The SVToR's production period lasts for 35 years. Figure 41: TOR and PSC regulatory framework and simultaneous production Buzios Case Source: PBR Libra The ANP estimates recoverable resources of 8-12bn boe with the lower end of the range representing Gaffney Cline's best estimate from 2010 (3.3-15bn boe on a % recovery factor), which at the time stated relatively minor concerns over trap and seal effectiveness, plus reservoir presence and quality. For Total and RDS, having taken a 20% stake each in the field for a price tag of $1.4bn would imply a finding cost of $ /boe not a bad price. Only one well had been drilled at the time of the bid round, and perhaps the lack of data on Libra and quality of seismic data in theory would make such an estimate on the resource base hard to believe. These are carbonate reservoirs (ie, can be highly heterogenous), we think that the ANP would have done the correct extrapolation to derive its figure. We note that 60 appraisal wells, 41 DSTs, 1.5km of core samples, 72 conventional logs and 57 production logs had been done in the pre-salt between 2006 and the time of the bid rounds. Big numbers that are crucial for PBR to keep enhancing its understanding of the pre-salt. The DoC deadline is a maximum of four years after contract signatures (or late 2017) with first oil required to be delivered within a five-year period following the DoC. The first large scale FPSO is targeted to be brought onstream in 2020 with many more thereafter, and RDS in Brazil 31

32 the pace and shape will depend on PBR's next Business Plan (September 2016). At the time of the bid round, the ANP mentioned a peak gross production potential of 1.4mbd. Ever since Libra was auctioned and awarded in October 2013, many signals continue to indicate this is a project that is being well managed and going well: Since December 2013 two months after the award PBR created a special 'Gerencia Executiva' (a separate organizational unit reporting directly to the Director of E&P) for Libra. The operational committee of Libra approved on time (end January) the 2014 investment plan for the block, and drilling activity started on time. The consortium is also well ahead to securing the FPSO for the extended well test in Libra, which should take place in The FPSO will have a capacity of 50kbd of oil and 4MM3/d of gas. Figure 42: Libra at the time of the auction Source: PBR; Note: CS assumes fewer FPSOs and lower end of the resource range at this time Sapinhoa (BM-S-9) PBR estimates 2.1bn boe of recoverable resources (up from an initial range of bn boe with the higher end at the time based on a 35% recovery factor) following the DoC in December 2011 (which happened one year ahead of schedule). The Guara discovery, now named Sapinhoa following the DoC, was announced in June 2008 and in 2009 the consortium re-entered the discovery well (Guara South) and performed a formation test, which showed that initial flow rates could reach up to 50kbd. At that time, most of the work had been carried out on the southern part of the field. It appears that Sapinhoa with its high productivity potential is not a candidate for CO2 injection due to its already high permo-porosity and strong water drive (aquifer). The appraisal well, which targeted the northern part of the field, showed an even better reservoir quality. BG s partner Repsol stated at the time that it had found a single-quality reservoir about 65m higher than the main discovery with a higher net pay. The second formation test on Guara North confirmed the high flow rate achieved in the southern part of the field. Sapinhoa was one of the first fields to be developed in the pre-salt Santos Basin, with the first pilot FPSO in early 2013 (120kbd oil capacity) followed by another FPSO (150kbd oil capacity because of lower carbon dioxide content in the area for the second FPSO) at the end of No further FPSOs are planned beyond the two already onstream for now. Lapa (BM-S-9) Lapa, formerly known as Carioca, is estimated to be ~460mmboe in size, according to PBR at the time of the DoC (BG stated that it sees further upside to this estimate). This RDS in Brazil 32