Vous êtes sur la page 1sur 33



Natalie Ballew, Kumar Das, Jeanne Eckhart, Stephen Hester, Reed Malin, John Maxwell, & Colin Meehan

Decision Pathways, Fall 2012

Abstract A premier hydrocarbon formation, Argentina's Vaca Muerta shale play has estimated recoverable reserves of 741 million barrels of oil and 4.5 trillion cubic feet of natural gas. Though a material contribution to any nation's petroleum reserves, Argentina's development of newly identified shale resources is especially important to offset, or even reverse, the nation's declining domestic energy production and debilitating energy trade imbalance. Growing reliance of fuel imports from Bolivia and Trinidad are the result of inadequate infrastructure investment, discouraging development policy, and low-set price controls. Due to trends toward nationalism and the recent expropriation of foreign assets, Argentina will struggle to exploit its shale resources independently and may be forced to leverage the expertise and capital of multinational firms. To analyze both the impact to the Argentine economy and potential profitability for multinationals, we determined the key variables involved in developing the Vaca Muerta and incorporated them into a system dynamics model. Outputs include net profits, jobs created, and tax revenue generated. Among the components used in the model's 24-year run-time are available geological findings from actual test wells in the region, extensive cost data from shale formations throughout the United States and Eastern Europe, current Argentine tax code and regulatory constraints, and commodity prices adjusted annually through probabilistic methods. Certain variables, such as the degree local and federal governments incentivize production and the number of wells drilled, are modifiable by the user. In spite of an upfront infrastructure cost of $500 million, the user's ability to manipulate several variables, and the complex interaction between the system's numerous random and looping variables, the model consistently yields multibillion dollar profits provided a reasonable number of wells drilled. Lack of data and the

inherent difficulty of predicting public policy limit the model's accuracy, but it serves as a useful tool in gauging the feasibility and challenges associated with developing the Vaca Muerta. Keywords: system dynamics, natural gas, energy policy

1. Introduction Argentina has a long history of gas use; in fact, a coal gas distribution system lit the city of Buenos Aires in 1856 (Maciel, 1992). Natural gas comprises nearly half of total energy consumption and about one-third of electrical power. For most of its history, Argentina was relatively self-sufficient when it came to the production of natural gas. However, domestically produced gas has declined 10% since peaking in 2006 as demand has risen by greater than 20% over the same period (EIA, 2012). From 2008 to 2011, the volume of imported liquid natural gas (LNG) increased by 900% at a price of $15 per million British thermal units (MMbtu), contributing to a large deficit in the energy trade balance (Gerold, 2012). In 2009, the worlds fourth largest shale play, the Vaca Muerta formation, was discovered in the Neuqun Basin in western Argentina. Development of these reserves has the potential to radically impact the nations economy, integration with global markets, and geopolitics. If fully developed, based on estimates by Exxon, Chevron, and YPF 1, oil and natural gas produced from the Vaca Muerta shale formation could meet Argentinas energy needs for over fifty years. Development, however, requires a seismic shift in national policy on multiple fronts and a multibillion-dollar infusion of capital and expertise from international firms (Gonzalez, 2012). In spring of 2012 Repsol, a Spanish oil and gas company with operations in Argentina, had its ownership stake in YPF partially seized (EIA, 2012) by the federal government. The official rationale behind the expropriation was to pressure other oil companies to grow production and penalize underinvestment (Ruano, 2012). International oil and gas firms play a critical role in developing and implementing advanced drilling and recovery techniques, in addition to providing the capital necessary for

Argentinas state-controlled energy company

development. It is unlikely Argentina would efficiently produce petroleum and natural gas from its shale formations without extensive involvement and support by foreign companies. Argentinas recent attitude toward international firms, such as Repsol, compromises the nations ability to secure vital expertise and capital. Preserving an equitable environment and upholding contracts with remaining energy companies is in sharp contrast to growing nationalistic tendencies. Outside of the failure to attract sufficient investment, extraction rates from the Vaca Muerta may fall short of expectations. Without long-term commitment from the Argentine people, development will be at constant risk. Despite massive production from the Barnett, Eagle Ford, and Marcellus in the United States (U.S.), extracting hydrocarbons from shale formations is relatively new technology with successful applications generally limited to the continental U.S. In contrast to conventional reservoirs, recovery from low permeability and porosity rock of shale formations has not been readily duplicated overseas (Carroll, 2012). In efforts to create a usable tool for firms considering investment in the Vaca Muerta, we use system dynamics modeling. Our model incorporates varying levels of foreign direct investment (FDI), tax burden, and other variables multinationals consider illustrating their impact on profitability. We first describe the past and present economic, political, and oil and gas environments in Argentina to gauge the value of developing of the Vaca Muerta. Next, we discuss the interactions, components, and structure of the system dynamics model created to analyze the impacts of development. We close with model results and general recommendations. 2. Background Development of the Vaca Muerta may have a profound effect on Argentine affairs. Recent economic progress has increased energy demand by 35% since 1998, the year domestic oil 4

production peaked. Declining reserves and increasingly arduous policy have led to Argentinas current multibillion-dollar annual energy trade imbalance. While Argentina remains the top producer of natural gas in South America, development of shale oil a viable solution to reversing the nations downward hydrocarbon production trend (EIA, 2012). However, the current political scene presents significant uncertainty to the future of development for the Vaca Muerta. 2.1. Politics Argentinas current political climate can best be described as turbulent. Since the financial collapse in 2000, Argentina has struggled to regain political and economic stability. The result has been increasingly nationalistic and protectionist economic policies. This has included increased restrictions on foreign direct investment and imports of many essential goods, which would be essential for any development. This section provides a brief overview of the political constraints that will affect any development in the Neuqun basin. 2.1.1. Regulatory Structure Argentina is a federal republic with 23 provinces; each province has its own set of regulatory statutes. All provinces have their own legal structure, linked and in compliance with the national constitution. From a development perspective, national projects such as infrastructure or tax structures are set at the national level, whereas labor and environmental regulations are a hybrid of local and federal agencies. Importantly, for exploitation of onshore resources, the individual provinces handle the concession and permit process. (Bravo, 2009) 2.1.2. Recent History Beginning in 1999 and continuing for nearly 3 years, Argentina experienced a complete financial collapse. The country experienced hyperinflation and ultimately defaulted on its entire foreign debtthe largest credit default in history. The legacy of this economic instability is still

highly present in Argentine politics. This extreme shift paved the way resurgence in nationalism and a return to protectionist trade and economic development policies. This resurgence is perhaps best understood through the political actors that emerged from this crisisNestor Kirchner and his wife Cristina Fernandez. Nestor Kirchner was elected to the presidential office in 2003 and oversaw the most significant part of the financial recovery. Under his tenure, the Argentine economy stabilized and began to add significant jobs in the industrial and agricultural sectors. Key to this was government support of Import Substituting Industrializationa strategy whereby strong barriers and taxes are placed on imports of specific goods in order to force purchases of local equivalents. Following his death in 2007, Nestor Kirchner was succeeded by his wife, who continued to implement similar economic strategies. Since 2010 the Fernandez government has more aggressively applied these tactics. In June of 2012, the government levied a new 14% tariff on all capital goods imported into the country (Morales and Castilla, 2012). This accompanies an already existing list of some 600 specific goodsincluding high-tech products like computers that have specific import restrictions. These actions have led to lawsuits against the Argentine government in the World Trade Organization. These protectionist policies have been implemented largely on the back of a wave of popular support and nationalistic political sentiments. The center-left leaning Fernandez has, for most of her first and second terms, enjoyed high approval ratings. She was easily reelected to a second term, however since her reelection she has faced public scrutiny as inflation continues to rise and many fear a repeat of the financial disaster of 2001. In the face of this increased scrutiny and pressureboth from within and abroadFernandez has stuck with her policies and driven an even harder nationalistic line. This has manifested itself through international sparring over

the contested Falkland Islands and most recently the re-nationalization of YPFthe former National Oil Company. This nationalist political discourse has had major impact on the growing Argentine energy sector. Though the Falklands conflict has its roots in the abortive armed struggle between the UK and Argentina, recent offshore oil discoveries have upped the stakes for political control of the island (Milmo, 2012). But most importantly the re-nationalization of YPF means that the Argentine state has directly invested itself in growing and promoting the countrys oil and gas reserves. This expropriationwhich was met with popular approval at home and nearly universal condemnation internationallywill ultimately place the responsibility and impetus for development of the Vaca Muerta field in the central governments hands. The Fernandez government now has the difficult task of enticing the same foreign investors that have been suffering under import restrictions to support YPF. 2.2. Economics Argentina historically produced oil and gas through state owned companies, YPF and Gas del Estado. In the 1980s and 1990s, modest reforms took place in the oil and gas industry. First, service contracts allowed upstream companies to either participate or pay commitment contracts. Between 1985 and 1990, Argentina allowed for a tax stabilization scheme during which 76 contracts were signed (Maciel, 1992). In 1992, the gas market was partially deregulated, leading to the privatization of Gas del Estado (Ennis, 2003). The natural monopoly of gas utilities was divided into four areas: production, local distribution, long-range distribution, and commercialization. The distribution systems were not privatized but production and commercialization were. A gas regulator, ENARGAS, was created to regulate the distribution

system and retail side of the gas market (Ennis, 2003, Ponzo et al. 2010). Repsol became the largest shareholder of YPF in the years after denationalization (Herrera and Garcia, 2003). This fragmentation negatively impacted gas market functionality as oil and gas production declined. Argentina previously prided itself on maintaining a robust domestic energy industry and for successfully exploiting its resources. From 1971 through 2004, Argentina increased overall production by about 725%. In 2004, natural gas production peaked at 1,892,272,637 MMBtu. From 2006 through 2011, production continued to fall and by 2011 was down over 10% from peak production. Natural gas demand continued to grow about 2% per year as GDP grew at 7% (IEA, 2012) 2. The aforementioned supply and demand imbalance was remedied through imports from Bolivia and Trinidad (EDI, 2012). Gas imports increased from 3,354,502 MMBtu in 2003 to 273,168,720 MMBtu in 2011, an increase of 800%. Amplified reliance on imports shifted the energy component of the trade balance negative; a major change from Argentinas historical precedent (Gerold, 2012). Besides electricity consumption, other top sectors of natural gas consumption are industrial and residential, which use 28% and 24%, respectively (EIA, 2012). Energy production is often a chief economic driver for countries endowed with substantial resources. The inverse also holds true; significant constraints on economic growth may occur if demand is met by costly fuel imports. Argentina was a net hydrocarbon exporter until 2004 when supply disruptions foreshadowed the extent of declining production. Inadequate natural gas supplies interrupted electricity generation and industrial sectors (Recalde, 2011). Proper development of the Vaca Muerta shale formation may lead to an increasing production of natural gas and economic stabilization.

In our model we assume a constant GDP. See Appendix C.

Figure 1: Argentina data from 1971 to 2010. From top left across then bottom left across: (a) GDP growth, (b) Population increase, (c) Natural gas production and imports vs. exports, and (d) Crude oil production and imports vs. exports (Data from EIA, 2012)

Another important aspect of the economic situation for the natural gas industry in Argentina is price controls initiated during the financial crisis in 2001 and 2002 (EIA, 2012, Ponzo et al. 2011). The policy was designed to put downward pressure on prices and alleviate the extreme inflation in the post-crisis period. The cap on gas prices was set at $2.50 per MMBtu and kept prices low in many areas. Following the initial success, however, the lack of incentives for producers caused debilitating gas shortages during peak demand periods. Winter and summer extremes usually coincided with insufficient supplies (EIA, 2012, Gonzalez, 2012). Though only initial development has begun in Vaca Muerta, there has been a great deal of speculation as to the potential economic impact of the field. The level of this impact will largely depend on the investment by foreign firms, government infrastructure build-up, and actual productivity of new wells. Because of this uncertainty, Argentina and its potential partners are looking to analogues in the U.S. where similar booms have occurred. Using metrics in relation to the Marcellus shale in Pennsylvania, the Vaca Muerta formation could have immediate impacts ranging in the tens of billions of dollars (Considine et al., 2011). 2.3. Oil and Gas The discovery of the Vaca Muerta formation significantly increased oil and natural gas reserves in Argentina. Estimates vary, but according to the Energy Information Administration (EIA), Argentina has approximately 2.5 billion barrels of oil in proven reserves, with the Neuqun basin containing 25%; this region is the second most productive oil province in Argentina. Figure 1 illustrates the vast natural gas resources in Argentina and the extent of the Neuqun Basin. The Vaca Muerta shale formation is predicted to have about 741 million barrels of recoverable oil. Proven natural gas reserves in Argentina are nearly 13.4 trillion cubic feet, despite a 50% decrease from ten years ago. 42% of proven and 50% of recoverable gas reserves


are in the Neuqun basin. With an estimated 4.5 trillion cubic feet of recoverable natural gas, the Vaca Muerta is among the most promising formations in Argentinas history (EIA, 2012).

Figure 2: Shale gas resources in Argentina (EIA ,2012)

2.4 Infrastructure: transportation, energy, and utilities Infrastructure and its expansion are essential for supporting hydrocarbon production. Road transport is the major means of transportation in Argentina. About 30% of Argentinas 200,000 km of roadways are paved and 80% of cargo is transported via roads. Recently, bids have been submitted for highway construction in the Neuqun province (BMI, 2012a). Rail transport is also a key element of Argentinas infrastructure. The nations length of operating track has declined from its peak shortly after World War II, but still serves over 500 million passengers annually. Swedish construction company, Skanska, is scheduled to complete a project 11

extending the railway line between Neuqun and Port of Bahia Blanca. (BMI, 2012a). 25.2 million tons or 12.5% of all freight was transported via rail in 2007. Modernization efforts are lowering costs, raising efficiencies, and increasing total rail capacity (INDEC, 2010). A wellfunctioning rail system will aid the movement of heavy machinery and personnel essential to develop the Vaca Muerta. In response to the adverse effect inadequate infrastructure and resource development has on its economy, the Argentinean government recently allocated $4 billion towards construction of gas pipelines and a re-gasification plant. Plans are also in place to build additional power plants and further diversify energy resources. These investments are beneficial but will not solve Argentinas larger problem of insufficient domestic energy production and the resulting energy trade imbalance (BMI, 2012b). The fluctuating presence of U.S. oil and gas majors ExxonMobil, EOG Resources, and Apache, and others has also alleviated infrastructure constraints (Morris, 2012). Apache plans to spend significant funds in 2012 for drilling and development of formations in Neuqun Basin. In recent developments, YPF and Chevron have reached an agreement to develop shale gas formations in Argentina. YPF plans to invest $7 billion annually through 2017 to increase hydrocarbon production. (BMI, 2012b). 3. Methodology To fully incorporate the complex interactions relevant to developing the Vaca Muerta formation, we use a system dynamics approach. The model seeks to understand the development of the Vaca Muerta over a twenty four-year period, incorporating variables pertaining not only to the formations geology and drilling parameters but also the political and economic environment.


To create a model we used Forio, an online software that allows users to create simulations using a system dynamics methodology. Forio allows input of variables, properties, and decisions using straightforward language. The end product is an online simulation through which players can explore the modeled system. This kind of simulation can aid firms and governments in determining whether development in the Vaca Muerta is worthwhile. First, each group member created an initial model of how we envisioned key variables interacting (see appendix for the drawn out images). We analyzed the individual models as a team to determine where each overlapped and what new interactions emerged from the collective. This method proved effective in considering the system from several perspectives given each member of the team differs in background. The interdisciplinary nature of our group, consisting of geologists, economists, and policy makers, was vital toward modeling the complexity of developing the Vaca Muerta.

Figure 3: Initial conceptual model progress

Next, we identified key variables and decisions and established the initial relationships of the model within Forio. Throughout the coding process we evaluated the logic and feasibility of outputs. From this iterative process we added micro-interactions or sub modules to create a more nuanced and dynamic model. Final primary components include the following: 13

TotalInfrastructureCost, TotalCost, TotalProfit, generated TaxRevenue, and ImportRegulation.3 Each consists of multiple sub modules; for instance, total profit interacts with nine other components.

Figure 4: Model interactions in Forio

3.1 Model Structure Forio allows iterative programming; we used one-year increments between 2012 and 2036. Key to system dynamics modeling, variables can store past calculations and engage in loops. Model behavior is structured in such a way that users can alter certain variables (decision variables, such as tax regime) prior to simulation. In addition, the number of wells (the determining factor in production rates and subsequently profitability, is modifiable at each step in the simulation. 3.2. Assumptions To create a manageable model and because of limited access and existence of information, we made assumptions concerning the formation geology, the oil and gas industry,

See Appendix A for more information on variables and decision variables used in the model.


the costs of development and production, and the policies of the Argentine government. All projections, however, are rooted in generally well-documented activities related to shale production and exploration throughout the United States and Poland. 4 It is not possible to cover all variables and associated assumptions in detail here, but we made significant efforts to deal with uncertainty and improve model reliability. For instance, projected commodity prices are treated as random variables constrained by the historically accurate commodity costs from data going back several decades. Over the 24-year period the model simulated, probabilistic approaches were implemented when possible. We determined this method optimal in dealing with uncertainty given the availability of sufficient historical data to calculate dependable figures. 3.2.1. Formation Geology For the simplicity of our model, we assume the Vaca Muerta formation is a uniform reservoir with no variability in production across different wells. Given the already limited information on the topic, our model does not account reserve growth in either oil or gas reserves in the Vaca Muerta; the life of the Vaca Muerta formation is complete when all estimated reserves are extracted. We assume annual declination rates of 77% and 89% (Swindell, 2012) for natural gas and associated liquids respectively. Rates change over time, but given the models time horizon, the impact on outcomes is minimal in comparison to the complexity of modeling more accurate and time-sensitive decline curves. Because of limited exploratory data, our model assumes no existing active wells in formation. Given the size of estimated reserves, the development of a few wells should not meaningfully alter the models results. For the wells we add in our model, we assume vertical
Poland has had recent experience with a similar production scenario (rural and undeveloped region) as the Vaca Muerta.


wells and horizontal wells produce the same amount of oil and gas. We found wells in shale formations across the U.S. produce 100 to 600 barrels of oil per day and 500 to 6000 MMBtu of natural gas. Therefore, oil production for wells in this model is 350 barrels per day and we agreed upon an equivalent gas capacity of 1,944 MMBtu per day (Seeley, 2012). The models time horizon and nature of its outputs allows us to assume oil and gas are extracted simultaneously. Annual well capacity estimation assumes 365 days in a year. In the code, estimated means recoverable reserves and liquids means only oil, not natural gas liquids. Figures within the model are estimations of recoverable oil and gas reserves; a significant increase in the recoverability factor or rate would necessitate minor model adjustments. 3.2.2. Costs The cost structure used in the model is vital to determine reliable profitability numbers and realizes several objectives. First, variables adjust to project size, which is achieved by focusing expenses down to the individual well level. This design allows model outputs, such as profitability, to adjust automatically if number of wells changes. Second, the cost structure incorporates available data while minimizing the impact of any individual presumption. Data comparable to what U.S.-based oil companies regularly provide on their operations is not widely available on Argentinas existing oil production. Specific cost estimates for developing the Vaca Muerta shale formation do not exist. Our approach was to leverage the significant volume of data on existing shale projects and fit it to the Vaca Muertas geology and geography while accounting for Argentinas infrastructure and domestic resources in the areas of heavy equipment, labor costs, and supply, as well as the pace of development. Activity in the Barnett, Eagle Ford, Marcellus, and other shale plays in the U.S. offer insight into well completion time frames, decreases in production costs over time, well pattern


density, infrastructure constraints and their effect on development, among other factors. Per well development costs began in the $8-13 million range and plateau at the $2.5-5 million depending on firm and location (Cowan, 2011). This array of data from years of quarterly earnings reports helped us identify the majority of cost variables and trends in how they changed over time. Announcements made by firms following the first series of wells drilled in an area revealed general consensus that drilling costs would decrease significantly, often by 50%, within the 24 months after initial drilling (EOG Resources, 2004). Crew and equipment relocation costs, road building, and economies of scale contribute to this phenomenon. Given many of these projections occurred in 2007 and 2008 as shale development accelerated, we were able to confirm these forecasts and establish a reliable cost decline curve. Within the code, primary cost variables include TotalInfrastructureCost, PerWellDrillingCost, and TotalDrillingCost. These variables are affected by diverse circumstances, such as GeneralInflationRate, TaxRegime, and an initial infrastructure cost of $500 million.5 Many parameters of Vaca Muerta formation development, however, cannot be accurately estimated using data from existing projects. For instance, there is no way to ascertain if logistical costs and expediency are comparable. Even relatively unpopulated regions of the continental U.S., such as North Dakota and Wyoming, have access to multiple interstate highways connected to the rest of the country. Specialized labor and most equipment is available domestically, if not locally. Argentinas comparably smaller oil sector would need infusions of foreign personnel and equipment to fully develop a shale play of Vaca Muertas magnitude. No comparable situation exists to help predict the logistical, and potentially more important, political complications multinational firms would likely encounter. Many of the same firms interested in the Vaca
Water is not considered a development cost or limitation to development; for our purposes we assume free and abundant water.


Muerta, such as ExxonMobil and Statoil, experimented in Polands recently discovered shale reserves. If Exxons experience in Poland is any indication, local population and transportation issues will be major hurdles (Carroll, 2011). Argentinas current political environment magnifies these risks considerably. Development costs averaged $11 million per well and upfront infrastructure costs are an estimated $500 million. In the model, we use this upfront costs based on the conditional similarities between the Polands shale play and the Vaca Muerta (Carroll, 2011). 3.2.4 Policies We assume that despite the vast area the Neuqun Basin covers, there will be consistent regulatory statutes across the region. Our model includes several policy choices regarding potential policy options at the federal and local levels. These options include tariffs on imported equipment, taxes on profits generated by foreign investments and local incentives to encourage local economic development resulting from well-drilling activities. Policy options are decisions made by the user. Import regulation impacts the cost of drilling and infrastructure as a tariff on imported equipment. Tax regime is a reflection of both the level of acceptance of foreign investment and domestic performances regarding resource protection; this influences tax revenue. Local tax incentives reflect local community willingness to support development; this directly affects per well drilling cost. Policy options are informed by existing regulation and current discussion regarding the development of natural gas in Argentina (Ernst and Young, 2010) In the case of tariffs on imported equipment the user may select a business as usual scenario which marginally increases the cost of drilling and infrastructure based on increased equipment costs; the decreased regulation scenario reduces this impact to a lesser degree; the


tariff free scenario prices drilling and infrastructure costs at the baseline costs discussed elsewhere in this paper. If the user chooses local incentives the per well drilling cost is reduced, reflecting the benefit of local incentives. 6 Another policy decision variable is federal tax policy, which may be seen as a reflection of both the level of federal acceptance of foreign investment and domestic preferences regarding resource protection. The tax rate is applied to total profit net of costs: the resource protection scenario sets the highest rate, business as usual sets a moderate rate and opportunity sets the rate at zero. These levels affect both the profitability of the venture and the domestic benefits of permitting foreign production. 4. Model development In an attempt to model the intricacies of this system we employed several of Forios mathematical operations. To assess the long-term viability of resource production from shale using decline curves, we used the stock function. This function enables the modeling of dynamic flows (i.e. flows that change over time) to address this challenge. The stock function is structured to begin with an initial value set by the user; in our model this value is the initial number of wells multiplied by the initial annual well output for gas and liquids. At each iteration of the model the user chooses whether to add additional wells and any additional wells produce at the expected initial year production capacity. Flows accumulate as wells are added, while production from old wells taper off quickly at the aforementioned rates, reflecting the realistic pattern of production from shale resources.

See Appendix A for policy option values.


The gas and liquids price in the model have been set to vary and float based on a stochastic formula with a random number generator. This change in gas prices seeks to model the variability that has existed within the Neuqun basin over the past ten year (Gerold, 2012). This model price assumption is also replicated within the liquids products as well. There is literature which supports the stochastic volatility to this type of modeling. As stated before, this model is simply expressed and hopes to approach a mean-regression model, which can be approximated over our model period. This stochastic behavior is also applied to domestic gas demand. 7 5. Results and Sensitivity Testing The purpose of this model is to assist in decision making around the exploitation of the Vaca Muerta Shale play and, as is the definition of a system dynamics model, there is no definitive result from these types of models. To account for the lack of tangible results we conducted a basic sensitivity test on a few potential common outcomes from running the model, trialing four basic scenarios: Business as Usual (BAU), BAU with Tax Incentives, All incentives, and Negative incentives. Each year it was assumed a static number of wells would be added, meaning a flat linear growth of total well capacity in Vaca Muerta. Also, incentives were held constant year-to-year, assuming a fixed policy path for development at the outset. This sensitivity testing revealed the following results on the total net profit, as described in Figure 4. With other factors held constant, Tax Incentives equal yielded the highest profit, more than combining incentives. This result is surprising, given that Free Trade incentives were perceived as being very important when constructing the model. Ultimately, raw taxes had a greater effect on total profit than combinations of alternative incentives.

See Appendix B for a more detailed description of using the stock function and stochastic price volatility.


Figure 4: Profit Scenarios Sensitivity Testing

6. Conclusion The models output is centered on overall drilling profit from all drilling companies on producing oil and natural gas from the Vaca Muerta shale formation. Aggregate profit is independent of the number of firms involved but is constrained by the estimated recoverable resources, a half-billion dollar up front infrastructure cost, taxes, foreign investment incentives, domestic demand, export potential, commodity prices, a maximum of wells that can be drilled during a given time period, and the rate of production, among other variables. The model proved flexible in exploring basic macro scenarios and in all cases returned positive profits; this is unsurprising given the nature of shale gas extraction and production. The strength of this model is its ability to test multiple scenarios on a yearly basisas the political situation in Argentina is in a constant state of flux. While the removal of a tax incentive or a trade barrier may have marginal impact (if programmed in at the beginning of analysis), if these incentives fluctuate from year-to-year the total profitand the total tax revenuewill be 21

affected. This is critical for international companies looking to invest and create risk profiles. The total economic impact of development of Vaca Muerta is a multi-faceted problem; the model presented here is a simplification for the purposes of identifying critical factors in development. The purpose of the model is not to be comprehensive, but rather to provide stakeholders with a baseline assessment of Vaca Muerta that highlights key economic indicators. Through use of this model, stakeholders and investors can begin to evaluate the value of the field and the varied approaches to its development available for selection.


Bogan, J., 2009. The Father of Shale Gas. Forbes Online. [Internet]. BMI, 2012a. Argentina: Industry Forecast Transport Q4 2012. Business Monitor International. [Internet]. Quarterly Report. BMI, 2012b. Argentina Infrastructure Report Q4 2012. Business Monitor International. [Internet]. Quarterly Report. Bravo, V. Sobre La Prorroga de las concesiones petroleras en la provincial de rio negro. Fundacin Patagonia tercer milenio. [Accessed 9 Dec 2012]. Carroll, J., 2012. Exxon Shale Failure in Poland May Lengthen Gazproms Shadow. Bloomberg. [Accessed 6 Dec 2012]. Considine, T., Watson, R., and Blumsack, S., 2011. The Pennsylvania Marcellus Natural Gas Industry: Status, Economic Impacts and Future Potential. The Pennsylvania State University, College of Earth and Mineral Sciences, Department of Energy and Mineral Engineering. [Internet] Cowen, T., 2011. Costs for Drilling the Eagle Ford. Rigzone. [Accessed 6 Dec 2012]. Deng, S., 2000. Stochastic Models of Energy Commodity Prices and Their Applications: Mean-reversion with Jumps and Spikes. University of California Energy Institute. Program on Workable Energy Regulation (POWER) Working Paper-073. EDI, 2012. Country Gas Profiles: Argentina. Energy Delta Institute. [Accessed 10 Dec 2012]. Available at <www.energydelta.org/mainmenu/energy-knowledge/country-gas-profiles/argentina>. EIA, 2012. "Argentina: Country Analysis Brief." Energy Information Administration. U.S. Department of Energy. [Accessed 28 Nov 2012]. Available at <http://www.eia.gov/countries/cab.cfm?fips=AR>. Ennis, H.M., Pinto, S.M., 2003. Privatization and Income Distribution in Argentina. The Effects of Privatization on Income Distribution in Latin America Project. [Accessed 5 Dec 2012]. EOG, 2004. EOG Resources reports first quarter 2004 results, increases 2004 production growth targets, reduces unit cost guidance and announces success in Texas Barnett Shale natural gas play. EOG Resources, Houston [Accessed 13 Dec 2012]. Press Release. Ernst & Young, 2010. Global Oil and Gas Tax Guide, 10-15. Ernst & Youngs Global Oil & Gas Center. Forio, 2012. About Forio. [Accessed 21 Nov 2012]. Available at <forio.com/about-forio/>.


Gerold, D.G., 2012. Argentina: E&P Business Status and Outlook. G&G Energy Consultants. Issue Brief. Gonzalez, P., 2012. YPF said to Expect Approval to Double Argentine Gas Prices. Bloomberg. [Accessed 25 Nov 2012]. Gonzalez, P., Klump, E., 2012. Exxon Joins Chevron in Weighing Argentina Shale Deals, YPF Says Bloomberg. [Accessed 6 Dec. 2012]. Herrera, C., Garcia, M., 2003. A 10 aos de la prvatizacin de YPF Anlisis y consecuencias en la Argentina y en la Cuenca del Golfo San Jorge. Centro Regional De Estudios Economicos De La Patagonia Central. [Accessed 28 Nov 2012]. IEA, 2012. International Energy Agency Data Services, Paris. [Accessed 29 Nov 2012]. Available at <www.iea.org>. INDEC, 2010. Instituto Nacional de Estadstica y Censos, Buenos Aires. Available in Spanish at <www.indec.mecon.fov.ar>. Maciel, R.N., 1992 Argentina: Privatisation of the natural gas industry. Energy & Nat. Resources 10, 371-79. HeinOnline. [Accessed 27 Nov 2012]. Milmo, D., 2012. UK explorers struggle to strike Falklands oil. The Guardian, London. [Accessed 9 Dec 2012]. Morales, M. and Castilla, J., Argentine president slaps tariffs on capital goods. Reuters. [Accessed 9 Dec 2012]. Orihuela, R., 2012. Argentina Seizes Oil Producer YPF, as Repsol Gets Ousted. Bloomberg. [Accessed 27 Nov 2012]. Ponzo, R., Dyner, I., Arango, S., Larsen, E., 2011. Regulation and Development of the Argentinean Gas Market. Energy Policy 39, 1070-079. [Accessed 27 Nov 2012]. Recalde, M., 2011. Energy Policy and Energy Market Performance: The Argentinean Case. Energy Policy 39, 3860868. [Accessed 26 Nov 2012]. Ruano, C., Stempel, J., 2012. Repsol sues Argentina over giant YPF seizure. Reuters. [Accessed 6 Dec 2012]. Seeley, R., 2012. YPF says Vaca Muerta shale could produce 300,000 b/d in 10 years. The Oil Daily. [Internet]. Schwartz, E.S., 1997. The Stochastic Behavior of Commodity Prices: Implication for Valuation and Hedging. The Journal of Finance LII (3), 923-973. Swindell, G.S., 2012. Eagle Ford Shale: An Early Look at Ultimate Recovery. Society of Petroleum Engineers. Presentation at the SPE Annual Technical Conference.


Appendix A: Table of key variables used in the model code Variable Name Type 8 Description Policy option selected by user, user selects different levels on tariffs of important drilling and pipeline equipment. Business as usual increases cost of drilling and infrastructure by 5%. Decreased regulation reduces this impact to 1.5%. Tariff free prices drilling and infrastructure costs at the baseline costs. Policy option selected by user, allows user to reduce the price of well drilling by 2% through local incentives. Policy option selected by user, includes stringent Resource Protection (30% tax), Business as Usual (15% tax) and Opportunity (0% tax). These levels affect both the profitability of the venture and the domestic benefits of permitting foreign production. Tax is imposed on all profits. Number of wells added each year by user Gas production decline rate for a single well Modeled stochastically, based on historical demand Total estimated resource, declines annually as production grows Price paid for exported gas Annual amount of gas produced, a function of production from new wells drilled in the current year and last year's production with the decline rate applied Inflator applied to per well drilling cost Cost to drill a single well A function of domestic and export revenues, drilling and infrastructure costs, with the selected TaxRegime applied PerWellDrillingCost*WellstoAdd Estimated cost of pipelines and other infrastructure needs to support oil and gas production Estimated production capacity of a gas well in the first year of production




WellstoAdd DeclineGasCurve DomesticGasDemand EstimatedGasResource ExportPrice GasProduction GeneralInflation PerWellDrillingCost TotalAfterTaxProfit TotalDrillingCost TotalInfrastructureCost WellGasCapacityYear


V indicates a variable. D indicates a decision variable, changeable by the user during model simulation.


Appendix B: Expanded description of applying STOCK function and floating commodity prices STOCK: The STOCK function is structured to begin with an initial value set by the user; in our model this value is the initial number of wells multiplied by the initial annual well output for gas and liquids. At each iteration of the model the user chooses whether to add additional wells or not; any additional wells produce at the expected initial year production capacity. Total production for that year is the combination of the initial or previous years production, and the production from wells added in the current year. Through the STOCK function the total production for the current year, n, is multiplied by the decline rate and used as the prior year production for the year n+1. As wells are added these flows accumulate, while production from old wells taper off quickly, reflecting the real life pattern of production from shale mineral resources.

Floating gas prices: The gas and liquids price in the model have been set to vary based on a stochastic formula with a random number generator. For example, in the model, the gas price is set at $5.2 per MMBtu, the price quoted in a G&G assessment for gas delivered to the at Neuqun province. The initial part of the formula is: V GasPriceInitial = 5.2. The next part of the price formula operates under the assumption of a forward price curve; the prices going forward are set to a percentage change in each year. This is given by: V GasPrice = NORMINV(RANDBETWEEN(0,1),1.2,.5)* GasPriceInitial. The mean price change is 20% with a standard deviation of 50%. This change in gas prices seeks to model the variability that has existed within the Neuqun basin over the past ten years (Gerold 2012). The price assumption is also replicated within the liquids products. Literature supports the stochastic volatility to this type of modeling. Schwartz (1997) and Deng (2002) present more complex models of stochastic volatility than our model. As Deng 26

states: several mean-reversion jump-diffusion modelsdescribe spot prices of energy commodities that may be very costly to store. I incorporate multiple jumps, regime-switching and stochastic volatility into these models in order to capture the salient features of energy commodity prices due to physical characteristics of energy commodities. This underlies aspects of our intent to set price initially and allow for volatility and variability in future projections of crude oil and natural gas prices. The mean regression is established in our model by incorporating a percentage change on a year-to-year basis rather than trying to project a direct price for several years into the future, which, with large standard deviations, would lead to negative prices. As stated before, this model is simply expressed and hopes to approach a meanregression model that can be approximated over the model time-span.


Appendix C: Model Code in Forio Language

########### Natural Gas Development Model ######## ################################### # Model properties M StartTime = 2012 M EndTime = 2036 M TimeStep = 1 M ExecuteDecisionImmediately = TRUE ################################### R Resource = Gas, Liquids #### Model Decisions #### D WellsStart = 10 V Initial NumberofWells = WellsStart D WellstoAdd = 0 V NoMoreWells = If(EstimatedGasResource <= 0, 1, 0) P NumberofWells.Label = Wells P NumberofWells.DecisionMin = 0 P NumberofWells.DecisionMax = 500 D DeclineGasCurve = 0.77 D DeclineLiquidsCurve = 0.89 D Price[Resource] = {2.5,8} V GasPriceInitial = 5.2 V GasPrice = NORMINV(RANDBETWEEN(0,1),1.2,.5)* GasPriceInitial P GasPrice.Label = USD V LiquidsPrice = NORMINV(RANDBETWEEN(0,1),8, 10) P LiquidsPrice.Documentation = www.forbes.com/2009/07/16/george-mitchellgas-business-energy-shale.html, P LiquidsPrice.Label = USD D ExportPrice[Resource] = {10,0} P ExportPrice.Label = USD www.slideshare.net/MarcellusDN/theeconomic-impact-of-the-value-chain-of-a-marcellus-shale-well

V NumberofWells = STOCK(WellstoAdd, Initial NumberofWells)

D TaxRegime = IF(ResourceProtection,


IF(BizUsualTax, IF(FIOpp,1,0), 0), 0) V ResourceProtection = 0.7 V BizUsualTax = 0.85 V FIOpp = 0.98 # This is intended to be a tax on all profits to reflect public policy D ImportRegulation = IF(TariffFree, IF(DecReg, 0), 0) V TariffFree = 1.00 V DecReg = 1.5 V BizUsualReg = 1.75 # This is a tax on imported equipment as reflected in the cost of infrastructure below D LocalTaxIncentive 0) V Incentives = 0.98 V NoIncentives = 1 # local incentives for localities to provide housing, etc.; incentive that only effects the perwellcost = IF(Incentives, I F(BizUsualReg,1,0),


D Initial DomesticGasDemand = 1600000000 D DomesticLiquidsDemand = 100000 #these numbers are in the right units (barrels/day) V RandomDemandFactor = RAND V DemandGrowth =NORMINV(RANDBETWEEN(0,1), 0.03, 0.08) V GasDemand = Initial DomesticGasDemand * DemandGrowth V LiquidDemand = DomesticLiquidsDemand * DemandGrowth #Randomizing demand

D GeneralInflationRate = 0.02 ### Input Variables ### V WellLiquidsCapacityDay = 350 * (WellstoAdd)


P WellLiquidsCapacityDay.Label = Barrels/day P WellLiquidsCapacityDay.Documentation = go.galegroup.com/ps/i.do?action= interpret&id=GALE%7CA282514322&v=2.1&u=txshracd2598&it=r&p=AONE&sw=w&auth Count V WellGasCapacityDay = 1944 * (WellstoAdd) P WellGasCapacityDay.Label = MMBtu/day V WellLiquidsCapacityYear = WellLiquidsCapacityDay * 365 P WellLiquidsCapacityYear.Label = Barrels/year V WellGasCapacityYear = WellGasCapacityDay * 365 P WellGasCapacityYear.Label = MMBtu/Year V DomesticGasDemand = STOCK(DemandGrowth*DomesticGasDemand, Initial DomesticG asDemand) #These functions determine the annual well output of gas & liquids based on daily output

V PerWellDrillingCost = STOCK(GeneralInflationRate*PerWellDrillingCost, Initl PerWellDrillingCost) * LocalTaxIncentive V Initial PerWellDrillingCost = 9000000 P PerWellDrillingCost.Label = USD P PerWellDrillingCost.Documentation = www.forbes.com/2009/07/16/georgemitchell-gas-business-energyshale.html, shale.typepad.com/haynesvilleshale/drillinginfo.drillinginfo.com/urb/barnett/, www.investopedia.com/sto costs/, ck-analysis/2009/drilling-and-completion-costs-continue-to-fall-in-shaleplays-swn-gxmrclr1002.aspx#axzz29pvwgbeq, www.slideshare.net/MarcellusDN/theeconomic-impact-of-the-value-chain-of-a-marcellus-shale-well V Initial EstimatedGasResource = 4500000000 P Initial EstimatedGasResource.Documentation = .cfm?fips=AR P Initial EstimatedGasResource.Label = MMBtu V EstimatedGasResource = STOCK(GasProduction, Initial EstimatedGasResource, NONNEGATIVE) P EstimatedGasResource.Label = MMBtu P EstimatedGasResource.Documentation = decline over time in gas resources


V Initial EstimatedLiquidsResource = 741000000 P Initial EstimatedLiquidsResource.DecisionMin = 0 P Initial EstimatedLiquidsResource.Documentation = www.eia.gov/countries/ cab.cfm?fips=AR


P Initial EstimatedLiquidsResource.Label = Barrels of oil V EstimatedLiquidsResource = STOCK(LiquidsProduction, Initial EstimatedLiquidsResource, NONNEGATIVE) P EstimatedLiquidsResource.Label = Barrels of Oil P EstimatedLiquidsResource.Documentation = decline over time in associate d liquids resources V Initial GasProduction = WellGasCapacityYear*WellsStart V GasProduction = IF(STOCK((DeclineGasCurve*GasProduction)+(WellstoAdd*WellGasCapacityYear), Initial GasP roduction, NONNEGATIVE)>EstimatedGasResource, EstimatedGasResource, STOCK((DeclineGasCurve*GasProduction) + (WellstoAdd*WellGasCapacityYear), Initial Ga sProduction, NONNEGATIVE)) V Initial LiquidsProduction = WellLiquidsCapacityYear*WellsStart V LiquidsProduction = IF(STOCK(DeclineLiquidsCurve*LiquidsProduction+WellstoAdd*WellLiquidsCapacityYear, Ini tial LiquidsProduction, NONNEGATIVE)>EstimatedLiquidsResource, Esti matedLiquidsResource, STOCK(DeclineLiquidsCurve*LiquidsProduction+WellstoAdd*WellLiquidsCapacityYear, Ini tial LiquidsProduction, NONNEGATIVE)) V TotalProfit = (TotalRevenue - TotalCost) P TotalProfit.Label = USD V TotalAfterTaxProfit = TotalProfit - TaxRevenue P TotalAfterTaxProfit.Label = USD V TotalRevenue = GasRevenue + LiquidsRevenue P TotalRevenue.Label = USD V Initial GasRevenue = 0 V GasRevenue = STOCK(GasProduction * GasPrice, Initial GasRevenue, NONNEGATIV E) P GasRevenue.Label = USD V PerWellGasOutput = (EstimatedGasResource) / NumberofWells P PerWellGasOutput.Label = MMBtu/well

V Initial LiquidsRevenue = 0 V LiquidsRevenue = STOCK((LiquidsProduction * LiquidsPrice) , Initial Liquids Revenue, NONNEGATIVE) P LiquidsRevenue.Label = USD V PerWellLiquidsOutput = (EstimatedLiquidsResource) / NumberofWells P PerWellLiquidsOutput.Label = Barrels of Oil/Year

V Initial ExportRevenue = 0 V ExportRevenue = STOCK((GasExportPotential * G 31

asPrice) + (LiquidsExportPotential * LiquidsPrice), Initial ExportRevenue, NO NNEGATIVE) P ExportRevenue.Label = USD V GasExportPotential = GasProduction - DomesticGasDemand V LiquidsExportPotential = TotalLiquidsProduction - DomesticLiquidsDemand V TotalCost = TotalInfrastructureCost + TotalDrillingCost P TotalCost.Label = USD V TotalInfrastructureCost = (500000000 + (200000 * NumberofWells)) * ImportRe gulation P TotalInfrastructureCost.Label = USD P TotalInfrastructureCost.Documentation = Houston Business Journal (Jun 30, 2 011) Report: Shale pipeline costs triple since 2004 V Initial TotalDrillingCost = 9000000*WellsStart V TotalDrillingCost = STOCK(PerWellDrillingCost * WellstoAdd, Initial TotalDr illingCost) P TotalDrillingCost.Label = USD V TotalLiquidsProduction = NumberofWells * WellLiquidsCapacityYear P TotalLiquidsProduction.Label = Barrels Oil/Year V TaxRevenue = (1 - TaxRegime) * TotalProfit P TaxRevenue.Label = USD V TaxRevGDP = TaxRevenue/445990000000 V DirectJobsPerWell = (NumberofWells * 11.95) V IndirectJobsPerWell = DirectJobsPerWell * 3.04 V TotalJobsPerWell = DirectJobsPerWell + IndirectJobsPerWell