Tuesday, December 11, 2007

Petrobras' Tupi Discovery Will Likely be Profitable

Petrobras announced the Tupi discovery in the Santos Oil Basin of between 5-8Bn barrels at 11.07, which appears to be recoverable numbers of oil and natural gas equivalent, as opposed to total resource numbers (of which recoverable oil is a fraction). The Tupi oil find generated significant enthusiasm both inside and outside of Brazil -- to the extent that the Brazilian President Luiz da Silva declared in a speech concerning the oil find that "God is Brazilian." Not only is the 5-8Bn barrels very large in itself, but, according to Petrobras CEO Gabrielli, Tupi is just a "tiny'' part of the total Santos Basin reserves.

The main questions that come to mind for the interested investor are: first, how much daily production can be expected from Tupi? And when will production come online? Further -- related to the first two questions -- how expensive will the Tupi field be to produce? And, further, what are the technical challenges to production? Overall, these questions can be grouped into an overall question: How profitable will Tupi be for Petrobras? This article will explore these questions, to the extent that the information has been made public as of the date of the writing of this article (12.07). Note that the author is not a petroleum geologist, so it is possible that mistakes will be found herein regarding technical issues -- the author has attempted to cite every assertion concerning technology issues with regards to the Tupi Oil Field.

Overall Significance of the Tupi Discovery to Petrobras:

As a perspective on the Tupi oil discovery, the latest oil production forecast for Brazil at 6.07 was presented by British Petroleum at 7/07 in the chart below. Note that this forecast most likely did not assume the successful production of Tupi specifically but rather some (most likely smaller) production success in the Santos Basin, as Tupi was not announced at 6.07 (and the Santos Basin was previously expected to produce mainly natural gas). From the forecast, one can summarize, Petrobras has several projects in the Campos Basin -- Tupi is located to the south in the Santos Basin -- which are expected to increase production regardless of how Tupi performs, in BP's estimate from approximately 2.0M bpd in 2007 to 4.0M bpd in 2020. Nevertheless, it is a significant positive if Tupi moves forward, from both a production and reserve standpoint -- Petrobras' 2P reserves were approximately 12 bn barrels and the Tupi discovery may move Petrobras' 2P reserves to between 17Bn to 20 Bn barrels (an amount significantly larger than Exxon Mobil's reported 1P reserves of 13.3Bn barrels at year end 2006). But in summary, Tupi is important, but not critical to the overall oil and natural gas production success of Petrobras.

Projected Future Oil Production of Brazil (pre-Tupi discovery):

Source: Historic, BP Statistical Review of World Energy, June 2007;
Forecast, BMI Research

How much oil will Tupi Produce?

Petrobras announced on 11/12/07 that Tupi production will go over 200,000 bpd in 10-15 years, with a pilot production of 100,000 bpd in 2011-2012. In terms of how much production can go over 200,000 bpd, the head of Petrobras exploration and production, Hugo Repsold, has indicated that 1M barrels per day of production is achievable, as a peak production figure.

The question of how much will be produced is related to questions concerning how the field will be produced from technological standpoint. If this field was a typical offshore oil discovery then the production numbers would likely offer little doubt as to their eventual achievability. But the main issue concerning Tupi is an unusually large and deep salt layer deep under the ocean floor, which covers the oil reservoir. A graphical depiction of this salt layer and the overall depth of the Tupi resource has been provide d by the excellent oil website The Oil Drum:

The salt layer combined with the large depths -- the Tupi oil discovery is 4 to 5 miles below the ocean floor -- make certain technologies necessary for the production of the field. The salt layer needs to be understood and effectively drilled through -- with a resulting stable wellbore -- requiring new technology and expertise to effectively manage through the salt layer. Salt at this depth is reported to act like sludge, with some unknown physical properties, meaning that construction of an effective wellbore may be difficult. Petrobras is assessed to have the technology to produce currently, as they have already drilled a number of test wells into the Tupi resource to make their size estimate public -- therefore the main question is cost. The field is likely to be expensive, which will be explored in the next section.

How Expensive will the Tupi Field be to Produce?

The respected consulting firm Wood MacKenzie has estimated that the total field will cost between $50 and $100Bn to produce, all in. Importantly, the cost range does not indicate the final production numbers from the field, so this makes a feasibility calculation (through a payback period, difficult to do (but a range under certain assumptions will be done in a section below, to shed light on Tupi). This amount comes in on top of Petrobras's planned $118Bn in spending announced for the next 5 years for all other projects, including refining, gas and ethanol pipelines, and other oil fields besides Tupi. These numbers indicate that Tupi will be expensive, but not prohibitively so for a firm of the size of Petrobras. But note that cost estimates for Tupi exhibit significant ranges. One Brazilian petroleum geologist has stated that costs for Tupi would be 10x higher than for oil fields produced in the Campos Basin, to the north, where the majority of Petrobras' current production exists. This would result in costs of over $100M per well. However, cost estimates have exhibited a very wide range, from $30M per well (according the the Petrobras CEO) to $250M per well -- the cost per well of the exploratory well. Full production of Tupi -- again Petrobras has not disclosed if this means 200,000 bpd or 1,000,000 bpd (which is a critical piece of information that is lacking currently) -- would require approximately 100 wells. Cost estimates for 100 wells in the Tupi field would then range from $3Bn at $30M per well to $25Bn at $250M per well.

Note that well costs tend to decline with more wells due to expertise and also the fact that some of the costs can be amortized over time -- the first well requires workers to be moved out to the location while more wells mean that labor and equipment are in place and ready to work translating to low transition costs. Therefore it is unlikely there won't be some cost abatement from the first test wells of $250M per well.

Further, there is a need for more Floating, Offloading and Storage Facilities (FOPF) -- which are offshore rigs but based on floating tankers instead of attached to the ground in order to produce the Tupi discovery. The costs of these FOPF's are in addition to the cost of drilling the wells. Petrobras' CEO Gabrielli has estimated that between 6 and 12 FPSO's will be needed to produce Tupi -- but did not indicate what amount of oil each FOPF would produce (in other words, how much production would 6 FOPF's produce verses 12 FOPF's?). There are only 70 FOPF's worldwide according to wikipedia at 2006, so likely shipbuilding facilities will be run overtime to provide several for the production of Tupi. -- with the bill of course going to Petrobras.

Estimate of Payback Period under Certain Assumptions:

Assuming an all in start up cost of between $50Bn and $100Bn - the Wood MacKenzie's estimate stated above - would mean a very wide ranging payback period of between 2 and 35 years, depending on the assumptions for production, oil price, and lifting costs. Note that lifting costs in the North Sea are estimated at under $15 per barrel of oil -- $20 is utilized for the cases below.

Scenario 1: Base Case:
500,000 bpd of total final production, oil price of $70, lifting costs of $20, total start up costs of $80Bn: Payback period: 7.3 years

Scenario 2: Low Case:
200,000 bpd of total final production, oil price of $60, lifting costs of $20, total start up costs of $100Bn: Payback period: 34.5 years

Scenario 3: High Case:
1,000,000 bpd of final production, oil price of $90, lifting costs of $20, total start up costs of $50Bn: Payback period: 2 years

*Note that in all cases, the time period in which capital expenditures are accrued but production has not come online is not included -- for example, if Petrobras takes two years to get to the point where some production is started up, these two years are not included in the payback period. Further, production numbers are assumed to be average numbers over the production period -- when in reality the field will rise to a "peak" then decline. Lastly Oil price is assumed to be constant over the payback period.

Clearly, the scenarios above present a very wide range, ranging from essentially uneconomic in the "Low Case Scenario" to massively profitable on the "High Case Scenario." The most likely case in the author's opinion -- named the base case above -- shows a 7.3 year payback period, which is at the upper limit for a payback period for a typical oil and gas project, as most oil firms would like to see a payback period of under 5 years, to reduce uncertainty and leave capital for other projects. However, for projects with a longer reserve life (larger reserves with more expected years of production), oil firms will likely be more willing to fund a longer payback period project. At 500,000 barrels per day of production, the Tupi oil field is expected to have a reserve life of between 27 and 47 years (at 5Bn and 8Bn barrels of reserves, respectively), which certainly fits the criteria for a long life asset, as, for example, most oil firms operating in the Gulf of Mexico have overall reserve lives of under 15 years.

The long life of the Tupi oil field is similar to oil sands projects, which have reserve lives of over 40 years. Oil sands firms are willing to fund projects with somewhat longer payback periods because of the long producing life -- the 5 year typical payback period is for conventional oil projects. For example, Suncor spent for the 4 year period through 2001, $3.4Bn in capital expenditures to double the capacity at its Millennium Project, which added an additional, approximate 130,000 barrels per day of production. In 2001, the oil price was significantly lower than it is today in late 2007, so it is likely Suncor was only budgeting between $10 and $15 per barrel in profit after the project was completed. These profit assumptions translate to a payback period of between 4.8 and 7.1 years (at $10 to $15 per barrel of profit) for Suncor's Millennium expansion.

As such, with long lived reserves but high initial capital expenditures, Tupi can be viewed similarly to a heavy oil project -- meaning that an expected pay back period of moderately over 5 years is acceptable and likely to be funded profitably.

In terms of the most likely cost and production scenario, the author would lean towards higher start up costs (closer towards the $100Bn estimate), but also a higher oil price going forward, and also somewhat higher than 200,000 barrels per day final production, with more than 6 FOPF's with 100 wells operating in the area. This scenario is reflected as mentioned in the "Base Case" above.

In summary, a cost estimate, roughly, shows that Tupi should be moderately profitable going forward, with risks of a lower realized oil price, cost overruns and lower realized production. These risks are partially offset by Petrobras' expertise in operating in deepwater -- Petrobras is the world's leader in offshore technology, and continued expected pressure on the oil price due to growth in Asian countries and lower expected production from conventional sources. Note that also there is significant upside to the profitability of Petrobras if the oil field is larger than initially reported.

Possible Benefits to Petrobras From Technological Expertise in Developing the Tupi Oil Field:

On the bright side, the technological expertise required to successfully extract the Tupi oil field will likely result in technology that allows for an enhanced ability to drill below deep salt layers, which will likely result in an increased ability to extract deep sea hydrocarbons in many other offshore areas with similar characteristics, such as West Africa and other South America. Further, preliminarily, Petrobras appears to be able to benefit from pioneering this deep sea salt drilling technology due to the ability to license and consult for other offshore projects with similar characteristics.


The Tupi oil find is quite promising, from a profitability standpoint, under most reasonable assumptions of oil price, cost and production figures. The expected profitability of Tupi is obviously quite positive for Petrobras investors. Further, Petrobras investors may see increased reserves and production from the Santos Basin beyond the initial 5-8 billion reserve estimate, which would boost profitability estimates. However, significant risks exist if oil prices drop going forward and/or the field proves to be too technologically challenging to develop, due to the unprecedentedly deep and thick salt layer. These risks are partially mitigated by Petrobras' expertise in deep water, and positive prospects for the price of oil. However, investors are encouraged to watch developments and news concerning the Tupi oil field closely over the next several years to determine if the technological risks are sufficiently mitigated.

No comments: