Kirby Wells, Wayne L. Kelley and Richard S. Bishop, RSK (UK) Limited, USA, examine the potential for cost overruns in Santos Basin pre-salt development and the potential strain on the operators financial capacity. Published in the October 2009 issue of Oilfield Technology.

Rightfully, much is made of the Brazilian Santos Basin pre-salt potential. Attention has been paid to the technical challenges of developing fields near the limits of water depth, distance from shore and sub-salt completion capability. There has also been interest in the costs of proposed technical solutions. The purpose of this article, however, is to examine the impact of the scale and rate of investment upon the capacity of the operators and financial markets to fund the development.

The authors will do this by determining a range of costs associated with development uncertainty, which grows with scale and complexity, and testing the range against an amalgamation of the current balance sheets of the Santos Basin operators.

A historical perspective
How accurate has historical costs forecasting been for mega-projects analogous in scale and complexity to the Santos Basin, and what is the range of potential cost outcomes for the development of the Santos Basin? Do the Santos BAsin operators have sufficient access to capital (retained earnings plus debt capacity) across the range of possible outcomes? The outhors look at these questions by using historical investment ratios as predictors for the operators’ investment behaviour projected over a decade long development period.

Prior to ’stress testing’ the financial statements of the operators, the authors looked at a range of possible development cost outcomes. Petrobras’ plans to invest US$ 111.4 billion in pre-salt development between this year and 2020 is assumed to be the current cost estimate and is used along with Petrobras’ 65% average working interest in the core Santos pre-salt to calculate a total, 10 year industry investment of US$ 171 billion. The authors then consider a sampling of pioneering oil and gas projects chosen for the availability of data, their complexity and their scale in terms of both volumes and investment. The projects are compared in a qualitative manner to illlustrate relationships among scale, complexity and accuracy of cost forecasts. The comparison is not intended to predict the likelihood or magnitude of potential outcomes. This historical perspective provides a range of outcomes that are more meaningful to discussion of Santos Basin developers’ financial capacity than any specific estimate.

Examining cost overruns
There are two primary contributors to cost overruns in mega-projects and neither is quantifiable by cost estimators. The first is a breakdown of the managerial process. Project management skill within the operator, service companies and engineering, procurement and construction (EPC) contractors is institutionalised at a elvel commensurate with normal activities. When unusually large demands necessitate an increase in management capacity, untested combinations of staff, new hiring and the creation of large project specifific management organisations dimish management efficiency. The second contributor is the inability to transfer risk at a large scale. Operators’ attemts to transfer development risk to service and EPC companies are only nominally successful. Should the financial burden become too great for the service or EPC companies to bear, the operator will provide relief, without regard to contract terms, before it will suffer more costly defaults, delays or shoddy performance. Inevitably, the operators as owners have a larger financial exposure to the project than the service providers. The operators’ larger capital base makes them the proverbial ‘deep pockets’

An attempt to examine cost overruns in large, complex projects quickly yields four insights:

1. There is too little information regarding cost estimates and ultimate outcomes.
2. Most of the information in the public domain consists of announcements of cost overruns or delays, and announcements heralding the completion of projects typically do not include a final tally.
3. While cubcontractors and operators proudly announce the on budget or near budget completion of a project, there is a noticeable absence of below budget announcements.
4. Allocation of costs is often opaquely misappropriated for a variety of reasons, making it difficult for third parties to determine true project costs.

These observations confirm the layman’s understanding of project cost estimates, that contrary to labels such as ‘expected’ and ‘most likely’, a project’s costs are much more likely to grow than to shrink. Given the scope of this article and scarcity of data, the authors prevail upon the reader to share this understanding as they perform an admittedly subjective review of cost forecasts for large, complex projects. The author’s firsthand experience in the development of the North Sea, Cantarell and the Tran Alaska Pipeline System (Alyeska) provided a unique opportunity to directly observe the strain and breakdown of operators’, service companies’, EPC contractors’ and fabricators’ management systems.

Selecting relevant projects
For the purpose of review, the authors selected projects including Sakhalin I and II, Alyeska, Cantarell, North Field GTL facilities, Thunderhorse, Forties and Brent. These projects include offshore fields, pipelines and facilities and are obviously not intended to be rigorous analogies to Santos Basin development. Nonetheless, the projects shared challenges to varying degrees resulting from the quality and quantity of geological and geophysical (G&G) and other early data that impact design assumptions, factors that increase development or execution complexity such as the need to develop new technologies or upscale existing technologies, the scale of the project in terms of its impact on the industry’s financial, manufacturing and technical resources, the volumes and recovery rates associated with the project, and the overall difficulty of managing the project. All of these attributes have the potential to independently increase cost and when combined they have a compounding impact. Figure 1 illustrates the resulting profile of several selected projects.

The contribution of scale and complexity
The authors’ intent is to illustrate the innately understood, but difficult to quantify, contribution of scale and complexity to cost overruns. It is an oversimplification to assert the complex projects are likely to exceed cost estimates and that small percentage overruns on large scale projects result in large magnitude overruns. The reality is that large scale, complex projects have a propensity toward large percentage cost overruns and long delays. The radar plots shown in Figure 1 demonstrate that, for all their differences, Alyeska and Santos Basin fields have a similar character when viewed in the light of those factors that led to staggering cost overruns in Alyeska. In this regard, the authors feel justified in using Alyeska as a high side sensitivity for the impact of multi-faceted challenges to a large scale project. Cantarell and the cancelled North Field GTL facilities demonstrate that cost overruns are possible even in projects that present much less complexity.
Another similarity between Santos development and others is the notion that a Santos development using scaled-up implementations of existing, recently proven technologies effectively limits the dangers of cost overruns or delays. In the case of Alyeska, there had been previous large diameter cross-country pipelines, pipelines that crossed permafrost and tundra, and as early as World War II product pipelines had been constructed across great distances in comparable conditions that brought refined goods from Canada to interior Alaska. Even with this experience, the cost estimate of upscaling this basic technology was remarkable in the magnitude of error. The original estimate, made in 1969, was US$900 million. The final cost without interest on loans in 1977 totaled US$8 billion; a cost overrun of nearly 800%. In the case of the North Sea, the same operators and contractors that had developed fields in the Southern and Central North Sea missed cost estimates by up to 48% when ’scaling’ the same facilities for the greater depth and wave heights in the Northern Sea.
Figure 2 compares the scale and complexity of some selected projects, as assigned by the authors, as well as an estimate of resulting cost overruns based upon publicly available data. It illustrates the magnitude of potential overruns in large, complex projects and the potential for scale and complexity to exacerbate each other’s impact on project cost.
Figure 3 shows a statistical distribution of project costs based upon the work of Emhjellen et al. in their 2001 study of North Sea development cost overruns. The authors added the relative outcomes of other large, complex projects and the identified capacities of current Santos operators and the industry as a whole.

Analysis of financial statements
The financial statements of the major operators with working interests in the Santos Basin have been analysed in Table 1 to estimate their collective financial capacity without regard for their respective proportional interests. Sonangol, Hydro and OGX have been excluded in the case of the first two companies there is inadequate public record information to determine their capacity and in the case of OGX there is insufficient history. Very small operators have been excluded because they make no statistical impact. The analysis presumes that all operators (with the exception of Petrobras) will limit borrowing to stay with their respective credit ratings. Credit ratings are a function of debt to equity ratios and the amount of forecast cash flow in excess of operating expenses and debt service. Balance sheets have been considered for their capacity to take additional debt and retaining current credit ratings without increasing shareholder capital. Cash flow is presumed to remain unchanged during the forecast period with no consideration to commodity price fluctuation. Financial reporting differs from company to company and a best estimate has been made to make comparisons with like attributes. Whatever error there may be in the interpretation of the financial statements to achieve parity for the magnitude of the analysis. Although it is a very large sum of money put at risk by one operator to develop a single project, Petrobra is taken at its word regarding its capacity to fund US$111.4 billion of the estimated us$171 billion development cost of Santos Basin pre-salt. Financial statements of the operators in the Santos Basin are analysed using the prior five years reporting included in Table 1.

Summary
The analysis of the amalgamated financial statements of the operators indicates:

1. There is sufficient combined capital capacity amongst the Santos Basin operators at the estimated cost for 10 year development of the Santos Basin pre-salt. However, cost overruns within the identified range of outcomes could require additional capitalization and/or redistribution of interests.
2. A cost overrun of approximately 125% (US$214 billion) will consume the available project capital. Cost overruns in excess of approximately 125% would require additional equity resources and/or reduced credit rating of the operators. Concurrent development of Tupi, Carioca, Jupiter, lara, etc. is probably not feasible.
3. Financial institutions are credited with estimates of as much as US$1 trillion of equity to support Santos Basin development and production over 20 years. It is doubtful the capital markets can provide the portion required in the next 10 years with the current market conditions and needs for funds in other industrial sectors if these predictions are accurate.
4. To develop the Santos Basin on schedule, there is a chance that cost overruns will require additional capitalization of the operators and/or sources of capital beyond the existing operators.