A LONG-TERM OUTLOOK FOR NATURAL GAS IN THE SOUTHERN …§ões/doutorado/Mauro... · 2021. 4....
215
A LONG-TERM OUTLOOK FOR NATURAL GAS IN THE SOUTHERN CONE: OUTCOMES FROM THE TIMES-CONOSUR MODEL Mauro Francisco Chávez Rodríguez Tese de Doutorado apresentada ao Programa de Pós-graduação em Planejamento Energético, COPPE, da Universidade Federal do Rio de Janeiro, como parte dos requisitos necessários à obtenção do título de Doutor em Planejamento Energético. Orientadores: Alexandre Salem Szklo André Frossard Pereira de Lucena Rio de Janeiro Setembro de 2016
Transcript of A LONG-TERM OUTLOOK FOR NATURAL GAS IN THE SOUTHERN …§ões/doutorado/Mauro... · 2021. 4....
A LONG-TERM OUTLOOK FOR NATURAL GAS IN THE SOUTHERN CONE:
OUTCOMES FROM THE TIMES-CONOSUR MODEL
Mauro Francisco Chávez Rodríguez
Tese de Doutorado apresentada ao Programa
de Pós-graduação em Planejamento
Energético, COPPE, da Universidade Federal
do Rio de Janeiro, como parte dos requisitos
necessários à obtenção do título de Doutor em
Planejamento Energético.
Orientadores: Alexandre Salem Szklo
André Frossard Pereira de Lucena
Rio de Janeiro
Setembro de 2016
iii
Chávez-Rodríguez, Mauro Francisco
A long-term outlook for natural gas in the Southern Cone:
outcomes from the TIMES-ConoSur model. / Mauro
Francisco Chávez Rodríguez – Rio de Janeiro: UFRJ/COPPE,
2016.
XVIII, 197 p.: il.; 29,7 cm.
Orientadores: Alexandre Salem Szklo
André Frossard Pereira de Lucena
Tese (Doutorado) – UFRJ/ COPPE/ Programa de
Planejamento Energético, 2016.
Referências Bibliográficas: p. 161-187.
1. Energy Planning. 2. Latin America. 3. Natural Gas
modelling. 4. Power generation. I. Szklo, Alexandre
Salem et al. II. Universidade Federal do Rio de Janeiro,
COPPE, Programa de Planejamento Energético. III.
Título.
iv
A todas las voces de la América Latina,
principalmente, a las menos escuchadas…
v
AGRADECIMENTOS
Tal vez sea ésta la parte más importante de toda la tesis, porque está relacionada
con otros seres humanos. Una tesis doctoral involucra muchos años de esfuerzo y
dedicación del autor, pero sobretodo refleja el apoyo de muchas personas por detrás. Es
por ello que escribo estos agradecimientos en el idioma que mejor sé expresar mi
gratitud.
Agradezco a mi familia, en especial a mi madre, Mirtha Rodriguez Montes de
Oca, que a pesar de la distancia, fue el soporte constante y la fuente de motivación de
mis empeños.
A mis orientadores y amigos, Alexandre Szklo y André Lucena, de quienes tanto
aprendí. Gracias por la manera con que practican el oficio de la enseñanza, y por me
haber dado el privilegio de compartir estos 4 años de trabajo con ustedes. Crecí mucho
como persona y profesional a su lado. Estaré eternamente agradecido.
A Roberto Schaeffer, que fue casi como un tercer orientador. Por lo aprendido en
los proyectos desarrollados en conjunto, las conversaciones y discusiones técnicas, y
por haber conducido, de alguna manera, a que el tema de esta tesis sea sobre gas natural.
A los miembros de la banca, Luiz Agusto Barroso, Helder Queiroz y Sergio
Bajay, por haber aceptado participar y prestigiar este trabajo. Su revisión y crítica será
fundamental para mejorar la calidad de este documento.
A Júlia Seixas, por la gentileza de haberme aceptado en su grupo de investigación
en mi viaje a Portugal, el CENSE, donde fue que aprendí TIMES. A la “maltinha gira”
del CENSE: Luis Dias (sei TIMES devido a você Luis), a Sofia Simões (quien bautizó
al modelo de TIMES-ConoSur), a João Pedro de Gouveia, Vera, Julihana, e Patrícia. A
los amigos de “Erasmus Lisboa”, especialmente a Ola y Andrea, por las fiestas en
Bairro Alto, los finales de semana de playa en Costa de Caparica, y muchas otras cosas.
A Adam Hawkes, por haber aceptado ser mi supervisor durante la estadía en el
Imperial College London, y por las discusiones sobre cómo modelar adecuadamente
proceso de la cadena de gas natural. A Johny Bosch, Francisca Jalil, Yukun, Cecilia,
vi
Vijay y Srikanth por su amistad y los cafés (y a veces pasteles) compartidos en mi
tiempo en Londres. A Antonella, por sus ganas de mejorar el mundo.
A Silvia Nebra, mi orientadora de maestría, cuyos consejos y amistad perduraron
en todo este tiempo.
A aquellos amigos con quien tuve la oportunidad de también ser co-autor en
publicaciones científicas: Rafael Garaffa, Pablo Carvajal, Alejandro Egüez, Juan
Martínez, Santiago Arango, Alexander Koberle, Daniela Varela, Fabiola Rodríguez,
Javier Bustos, Ricardo Raineri, Gerardo Rabinovich, Gisela Andrade y Gónzalo
Cárdenas. Con ustedes verifiqué que elaborar un paper es más emocionante y placentero
cuando se trabaja en equipo.
A las personas que fueron una “familia” en Río de Janeiro: Rafael Soria (qué
honra haber sido padrino de tu matrimonio), Esperanza Gonzalez, Joana Portugal,
Camilla Oliveira, Eveline Vásquez, Lina Paz y Catalina Gavilanes.
A los amigos de CENERGIA, por todas las alegrías: Raúl, Paulo, David, Susi,
Bruno, Fernanda, Mariana, Ana Luiza, Tamara, Regis, Larissa, Belinha, Diego, Cindy,
Pedro. A los Yuvas del Sahaja Yoga: Sthela, João, Marina, Rafael, Aristides, Monique.
A Leoncio, por el soporte y conocimiento recibido.
A las personas que hicieron más feliz estos 4 años de doctorado, llenando mi
mente de memorias gratas: Luisa, Daniela, Sylvia, Naila, Daniel Howard, Michele,
Judith, Renato, Priscilla, Reynaldo, Lissel, Ricardo, Maricarmen, Ana, Elby, José Koc,
Luis Haro, Nella, Olivia, Shirley, Niagara, Guilherme y Denise.
Agradezco también a Wood Mackenzie, por la oportunidad de trabajar con
excelentes profesionales y permitirme continuar desarrollando las actividades que más
me apasionan en mi carrera.
Y finalmente, al apoyo financiero del CNPq-TWAS que hicieron todo esto
posible.
vii
Resumo da Tese apresentada à COPPE/UFRJ como parte dos requisitos necessários
para a obtenção do grau de Doutor em Ciências (D.Sc.)
UMA PROJEÇÃO DE LONGO PRAZO PARA O GÁS NATURAL NO CONE SUL:
RESULTADOS DO MODELO TIMES-CONOSUR
Mauro Francisco Chávez Rodríguez
Setembro/2016
Orientadores: Alexandre Salem Szklo
André Frossard Pereira de Lucena
Programa: Planejamento Energético
Esta tese foca no desenvolvimento de um modelo de gás natural para analisar o
papel de fatores chave que vão determinar a futura dinâmica dos mercados de gás
natural no Cone Sul. O TIMES-ConoSur foi testado para a avaliação do balanço da
oferta e demanda de gás desde 2012 até 2030 em dois cenários: O “Constrained
Scenario”, onde o CAPEX do upstream é restringido na Argentina e o gás associado dos
campos offshore no Brasil seguem uma curva de produção conservadora; e o
“Unconstrained Scenario”, sem limites financeiros na Argentina e uma curva de
produção otimista de gás associado Brasileiro. Os resultados mostram que no
“Unconstrained Scenario” as importações de GNL acumuladas reduziriam em 54% e o
comércio regional incrementaria em 15%, quando comparados ao “Constrained
Scenario”. O gás não convencional da Argentina poderia mudar o jogo, com o potencial
para reduzir a menos da metade as importações de GNL na Argentina, retomar as
exportações para o Chile, Brasil e ainda aos mercados internacionais, se os preços de
GNL subirem. O gás associado Brasileiro contribuiu com a autossuficiência de gás do
país, porém o fornecimento de gás Boliviano ainda é a solução de menor custo. As
tecnologias solar e eólica reduzem a parcela do mercado do gás na geração elétrica,
porém fortalecem a relevância do gás como uma fonte confiável de respaldo. TIMES-
ConoSur é um modelo integrado de gás pioneiro para propósitos de planejamento na
América Latina.
viii
Abstract of Thesis presented to COPPE/UFRJ as a partial fulfillment of the
requirements for the degree of Doctor of Science (D.Sc.)
A LONG-TERM OUTLOOK FOR NATURAL GAS IN THE SOUTHERN CONE:
OUTCOMES FROM THE TIMES-CONOSUR MODEL
Mauro Francisco Chávez Rodríguez
September/2016
Advisors: Alexandre Salem Szklo
André Frossard Pereira de Lucena
Department: Energy Planning
This thesis aims at developing a natural gas model, TIMES-ConoSur, to assess the
role of key factors that will determine the future natural gas markets dynamics in the
Southern Cone. TIMES-ConoSur was tested for evaluating the natural gas supply and
demand balance from 2012 to 2030 under two scenarios: the Constrained Scenario
where upstream CAPEX is restricted in Argentina and associated gas from offshore
fields in Brazil follows a conservative supply curve; and the Unconstrained Scenario,
with no financial limits for Argentina and an optimistic Brazilian associated gas
production curve. Findings show that in the Unconstrained Scenario accumulated LNG
imports reduced in 54% and intra-regional gas trade increased in 15%, when compared
to the Constrained Scenario. Unconventional gas in Argentina can be a game-changer,
with the potential to reduce by less than a half the LNG imports in Argentina, retake gas
exports to Chile, Brazil and even for the international market, if LNG prices rise.
Brazilian associated gas contributed with the gas self-sufficiency of the country,
however Bolivian natural gas supply remains the least-cost solution of the model. Solar
and wind power reduce the natural gas market share in electricity generation, but
strengths its relevance as a backup reliable source. TIMES-ConoSur is a pioneer gas
focused integrated model developed for energy planning purposes in Latin America.
ix
INDEX
1 INTRODUCTION .................................................................................................... 1
1.1 The relevance of natural gas ................................................................................. 1
1.2 Natural gas in the Southern Cone ......................................................................... 3
1.3 Research Question, Aims and Objectives............................................................. 8
1.4 Overview .............................................................................................................. 9
2 LITERATURE REVIEW ...................................................................................... 10
2.1 Natural gas models ............................................................................................. 10
2.2 Southern Cone Gas Market Studies .................................................................... 17
2.3 Key factors for the Southern Cone natural gas market ....................................... 21
2.3.1 Argentina’s hydrocarbon industry and its unconventional resources.......... 21
2.3.2 Petroleum production in Brazil, the offshore associated gas, and the
Amazonian resources .............................................................................................. 23
2.3.3 Regional trade of natural gas: the Argentina-Chile case and the Bolivian
exports .................................................................................................................... 27
2.3.4 The LNG expansion in the Southern Cone ................................................... 30
2.3.5 The Hydropower in Brazil and Wind and Solar potential in the Southern
Cone 33
3 METHODOLOGICAL PROCEDURE ................................................................ 37
3.1 Research Design ................................................................................................. 37
3.1.1 Concept ......................................................................................................... 37
3.1.2 Modelling Approach ..................................................................................... 41
3.1.3 Geographical Resolution .............................................................................. 44
3.1.4 TemporalResolution ..................................................................................... 45
3.2 Tools description ................................................................................................ 47
3.2.1 LEAP ............................................................................................................. 47
3.2.2 TIMES ........................................................................................................... 48
4 NATURAL GAS DEMAND FOR END-USES MODELLING .......................... 53
4.1.1 Residential sector ......................................................................................... 55
4.1.2 Commercial and Public Sectors ................................................................... 59
4.1.3 Industrial Sector ........................................................................................... 60
4.1.4 Transport Sector ........................................................................................... 63
x
5 NATURAL GAS SUPPLY TECHNOLOGIES IN TIMES-CONOSUR .......... 66
5.1 Power Generation Modelling ............................................................................. 69
5.2 Upstream Modelling ........................................................................................... 76
5.2.1 Upstream natural gas consumption and losses ............................................ 79
5.2.2 Non-Associated Natural Gas Production Modelling ................................... 80
5.2.3 Associated Natural Gas Production Modelling ........................................... 85
5.2.4 Associated Natural Gas Options .................................................................. 90
5.2.5 Reserves, Resources and Costs ..................................................................... 92
5.3 Midstream Modelling ....................................................................................... 100
5.3.1 Natural Gas Processing Plants .................................................................. 100
5.3.2 Regasification and Liquefaction of Liquefied Natural Gas(LNG) ............. 103
5.3.3 Gas Pipelines .............................................................................................. 107
5.4 Scenarios ........................................................................................................... 110
5.4.1 Investment capacity in upstream as the main constraint for the future natural
gas market in the Southern Cone .......................................................................... 112
5.4.2 Constrained Investment Scenario ............................................................... 115
5.4.3 Unconstrained Investment Scenario ........................................................... 116
6 RESULTS .............................................................................................................. 118
6.1 Natural gas domestic production results ........................................................... 118
6.2 Natural gas supply results ................................................................................. 122
6.3 Power Generation Results ................................................................................ 125
6.4 Natural Gas Demand ........................................................................................ 129
7 DISCUSSIONS ..................................................................................................... 135
7.1 What is the economic potential ofArgentina’s unconventional? ...................... 135
7.2 What role LNG will play in the natural gas market of the Southern Cone?..... 138
7.3 Will Brazil be gas self-sufficient using its associated gas production from
offshore fields? ......................................................................................................... 141
7.4 Will regional natural gas trade in the Southern Cone increase with the current
pipeline infrastructure? ............................................................................................. 144
7.5 Will renewable energy affect the natural gas consumption in power generation?
149
8 FINAL REMARKS .............................................................................................. 154
9 REFERÊNCIAS BIBLIOGRÁFICAS ............................................................... 161
xi
10 ANNEX .................................................................................................................. 188
10.1 ANNEX A: NATURAL GAS DEMAND SUPPLEMENTARY MATERIAL188
Natural gas demand split in North Chile and Central-South Chile .......................... 188
Econometric approach for the natural gas demand projection in industrial sector .. 190
10.2 ANNEX B: MODELLING SUPPLEMENTARY MATERIAL ...................... 193
Power Generation Technologies ............................................................................... 193
Multi-Hubbert parameters ........................................................................................ 194
Gas-to-oil ratio for North Brazil ............................................................................... 195
Natural Gas reinjection in Oil Fields ........................................................................ 196
xii
INDEX OF FIGURES
Figure 1.1. Evolution of primary energy consumption in the four studied SC countries.
Source: CNE (2016a); EPE (2016a); MHE (2015); SECRETARIA DE ENERGIA
(2015a) .............................................................................................................................. 4
Figure 1.2. Natural Gas consumption structure by sectors in the Southern Cone in 2014.
Source: CNE (2016a); EPE (2015a); MHE (2015); SECRETARIA DE ENERGIA
(2015a). ............................................................................................................................. 5
Figure 1.3. Natural gas net production and imports in the Southern Cone countries.
Source: Own elaboration based on ANP (2007, 2015); BP (2016); CNE (2016); EIA
(2015); INE (2014b); SECRETARIA DE ENERGIA (2015a). ....................................... 6
Figure 1.4. Natural Gas Transport and Regasification Infrastructure in the Southern
Cone. ................................................................................................................................. 7
Figure 2.1. Shale resources in Argentina. ....................................................................... 23
Figure 2.2. Top natural gas production fields in Brazil in January, 2016. Source: (MME,
2016b) ............................................................................................................................. 25
Figure 2.3. Historical natural gas consumption in Chile and Argentina in Buildings
(residential, commercial and public sector) and power generation sectors between 2006
and 2011. Source: CNE (2016c) and ENARGAS (2015) .............................................. 31
Figure 2.4. A ship-to-ship transfer of LNG in Bahia Blanca Gas Port........................... 32
Figure 2.5. Map of Brazilian Amazon showing existing and planned hydroelectric
power plants. Source: ALMEIDA PRADO JR. et al. (2016) ......................................... 34
Figure 2.6. Historical capacity factor between 2012 and 2015 for hydropower plants in
Brazil. Source: ONS (2016)............................................................................................ 35
Figure 2.7. Investment suitability for (a) Wind power and (b) Solar power in South
America. Source: IRENA (2016) ................................................................................... 36
Figure 3.1. Equilibrium of the supply and an elastic demand of a specific commodity. 38
Figure 3.2. Equilibrium of the supply and a fixed demand of a specific commodity. ... 39
Figure 3.3. Natural Gas & Power modelling approach developed for this study ........... 42
Figure 3.4. Geographical Resolution of the TIMES-ConoSur model ............................ 45
Figure 3.5. Example of LEAP’s interface. Source: HEAPS (2012) ............................... 48
Figure 3.6. Schematic of TIMES inputs and outputs. Source: REMME et al. (2001) ... 51
xiii
Figure 4.1. Methodological procedure for the elaboration of natural gas demands for
end-uses. ......................................................................................................................... 54
Figure 4.2. Natural gas consumption curve estimated for households by end use for
Chile and Argentina. Source: Own elaboration based on ENARGAS (2015) and CNE
(2016c). ........................................................................................................................... 57
Figure 4.3. Projected vehicle ownership until 2030. ...................................................... 64
Figure 5.1. TIMES-ConoSur architecture ...................................................................... 67
Figure 5.2. Power Generation Capacity Expansion by technologies. ............................ 71
Figure 5.3. Commodity Fraction of the electricity demand modelled in the base year. 72
Figure 5.4.Monthly capacity factors adopted for Hydropower plants by countries. ...... 72
Figure 5.5. Monthly capacity factors adopted for wind power plants by countries. ...... 73
Figure 5.6. Hourly capacity factors adopted for solar power plants by countries (PV
solar power). Source: Based on CNDC (2016); CNE (2016b); SECRETARIA DE
ENERGIA (2014); SORIA (2016) ................................................................................. 73
Figure 5.7. Fuel Prices assumed for power generation. Source: Based on ANH (2015);
ANP (2015); CAMMESA (2015); CNDC (2013); CNE (2015); EPE (2016a). ............ 74
Figure 5.8. Condensates to Wet Gas ratio on energy basis assumed for non-associated
gas for each region. Source: SGI (2016) ........................................................................ 77
Figure 5.9. Natural gas production profile of a typical field. ......................................... 83
Figure 5.10. Estimated natural gas production requirement and natural gas production
capacity curves according to reserves/resources classification, an example of Bolivia.
Source: CHAVEZ-RODRIGUEZ et al. (2016a) ............................................................ 84
Figure 5.11. Multi-Hubbert Curve for oil production calculated for Argentina. ............ 87
Figure 5.12. Historical natural gas to oil ratio from oil fields in Post-Salt and Pre-Salt in
Brazil. Based on MME (2016a) ...................................................................................... 88
Figure 5.13. Oil production in the North Region projected to 2030. Based on: ANP
(2015) ............................................................................................................................. 89
Figure 5.14. Natural Gas Separation and Processing in the FPSO. Source: DE MORAES
CRUZ et al. (2016) ......................................................................................................... 91
Figure 5.15. Low-Temperature Fischer-Tropsch Process Overview. Source: SBM
OFFSHORE (2014) ........................................................................................................ 92
Figure 5.16. Discounting at present value the CAPEX procedure for the inputting cost in
the model. ....................................................................................................................... 98
xiv
Figure 5.17. Natural Gas Processing Plant Schematic. ................................................ 101
Figure 5.18.CAPEX estimations vs. Capacities of different Natural Gas Processing
Plants. ........................................................................................................................... 102
Figure 5.19. Diagram of a natural gas liquefaction process. ........................................ 104
Figure 5.20. LNG Regasification Schematic Process. ................................................. 105
Figure 5.21. Average prices of LNG imports in the Southern Cone. ........................... 107
Figure 5.22. Investments in upstream CAPEX per year in Argentina considered for the
Constrained Investments Scenario. .............................................................................. 115
Figure 5.23. Brazil’s oil production curve for the Constraint Investment Scenario. .... 116
Figure 5.24. Brazil’s oil production curve for the Unconstrained Investment Scenario.
...................................................................................................................................... 117
Figure 6.1. Gross natural gas domestic production under the Constrained Investments
Scenario ........................................................................................................................ 119
Figure 6.2. Gross natural gas domestic production under the Unconstrained Investment
Scenario ........................................................................................................................ 120
Figure 6.3. Natural gas supply projections in the Southern Cone under the Constrained
Investment Scenarios .................................................................................................... 123
Figure 6.4. Natural gas supply projections in the Southern Cone under the
Unconstrained Investment Scenarios ........................................................................... 124
Figure 6.5. Power generation profiles in the Southern Cone under the Constrained
Investment Scenario ..................................................................................................... 127
Figure 6.6. Power generation profiles in the Southern Cone under the Unconstrained
Investment Scenario ..................................................................................................... 128
Figure 6.7. Natural gas demand projection in the Southern Cone under the Constrained
Investment Scenarios .................................................................................................... 131
Figure 6.8. Natural gas demand projection in the Southern Cone under the
Unconstrained Investment Scenarios ........................................................................... 132
Figure 7.1. Unconventional non-associated gas production in Argentina under the
different scenarios assessed .......................................................................................... 136
Figure 7.2. Levelized cost of Argentinian non-associated gas production used in
TIMES-ConoSur. .......................................................................................................... 137
Figure 7.3. Average Well Drilling and Completion Cost for some Unconventional
Basins in the USA. Source: EIA (2016c) ..................................................................... 138
xv
Figure 7.4. Accumulated imports (+)/exports (-) of LNG from 2016 until 2030 for
different LNG price changes. ....................................................................................... 140
Figure 7.5. Natural gas supply in Brazil in a self-sufficient sensitivity scenario. ........ 142
Figure 7.6. Gross natural gas production in Brazil in the self-sufficient sensitivity
scenario. ........................................................................................................................ 143
Figure 7.7. Natural gas trade between Bolivia and Brazil under different scenarios ... 145
Figure 7.8. Natural gas trade between Bolivia and Argentina under different scenarios
...................................................................................................................................... 145
Figure 7.9. Natural gas trade between Argentina and Chile under different scenarios. 147
Figure 7.10. Natural gas exports from Argentina to Brazil under different scenarios. 148
Figure 7.11.Natural gas supply sensitivity scenario for Brazil excluding São Luiz de
Tapajos hydroelectric project and with drought years.................................................. 150
Figure 7.12. Power generation simulated by sources in the Southern Cone in 2015 and
2030 .............................................................................................................................. 151
Figure 7.13. Power operation in Chile simulated for 2030. ......................................... 152
Figure 10.1. Energy consumption structure projected (from 2013 onwards) by fuels
excluding electricity in Chile’’s industrial sector. ........................................................ 191
Figure 10.2. Forecasted useful energy-excluding electricity- in the Chilean industrial
sector. Source: (IEA, 2014b). ....................................................................................... 192
Figure 10.3. /oil produced
ratio equation adopted for the North Region. ............................................................... 196
xvi
INDEX OF TABLES
Table 2.1. Review of natural gas modeling in Energy Planning .................................... 11
Table 4.1.Share of households for the cooking energy service. ..................................... 58
Table 5.1. Acronyms description used in Figure 5.1 for process technologies .............. 68
Table 5.2. Interconnection capacity for domestic energy trade considered between
regions in Brazil and Chile. Source: Based on E-CL (2015); EPE (2015b) ................... 75
Table 5.3. Natural gas classification according to gas-to-oil volumetric ratio (v/v) ...... 76
Table 5.4. Consumption and Losses parameters considered for upstream modelling ... 79
Table 5.5. EUR in Mm3 of natural gas used in each region........................................... 96
Table 5.6. Non-associated natural gas production costs adopted for the modelling. ..... 99
Table 5.7. Yields adopted to model the fractionation process...................................... 103
Table 5.8. Regasification Terminals in the Southern Cone. ......................................... 106
Table 5.9. International natural gas pipelines in the Southern Cone. ........................... 109
Table 5.10. Summary of the main scenarios simulated in this thesis ........................... 111
Table 6.1. Natural Gas Balance modeled in the Southern Cone for different Scenarios
...................................................................................................................................... 134
Table 7.1. Total system discounted cost (MUS$) for the different LNG prices
sensitivities ................................................................................................................... 141
Table 7.2. Research questions and answers provided by TIMES-ConoSur ................. 153
Table 10.1.Sectorial natural gas consumption in Mm3 in Chile for 2006 and 2012. ... 189
Table 10.2. Natural gas demand Split between Chile regions for base-year and future
increments of demand. .................................................................................................. 190
Table 10.3. ARIMA model for the natural gas consumption in the Argentinian industrial
sector. ............................................................................................................................ 190
Table 10.4. VEC model parameters for natural gas consumption in industrial sector and
GDP .............................................................................................................................. 191
Table 10.5. VEC model for useful energy consumption in industrial sector –excluding
electricity- and industrial GDP of Chile. ..................................................................... 193
Table 10.6. Techno-Economic parameters for the power generation Technologies
modelling. ..................................................................................................................... 194
Table 10.7. Multi-Hubbert model parameters used for Argentina. .............................. 195
Table 10.8. Multi-Hubbert model parameters used for Brazil. .................................... 195
xvii
ACRONYMS
ARIMA Autoregressive Integrated Moving Average
bcm Billion cubic meters
boe Barrels of oil equivalent
Btu British Thermal Units
BRAZIL-GIN Integrated Brazil
BRAZIL-NR North Brazil
CAPEX Capital Expenditure
CCS Carbon, Capture and Storage
CNG Compressed Natural Gas
CO Carbon monoxide
CO2 Carbon Dioxide
EFOM Energy Flow Optimisation Model
EPE Energy Research Company (Empresa de Pesquisa Energética)
EU European Union
EUR Estimated Ultimate Recovery
ETSAP Energy Technology Systems Analysis Programme
GasBol The Bolivia–Brazil pipeline(Gasoduto Bolívia-Brasil)
GDP Gross Domestic Product
GHG Greenhouse Gases
GTL Gas to Liquids
IEA International Energy Agency
LEAP Long range Energy Alternatives Planning System
LNG Liquefied Natural Gas
LPG Liquefied Petroleum Gas
MARKAL The MARKet ALlocation model
Mm3 Million standard cubic meter
NGL Natural Gas Liquids
NGPP Natural Gas Processing Plants
NOC National Oil Companies
OPEX Operational Expenditure
xviii
PEMAT Plan for the Gas Pipeline Transport Network (Plano Decenal de
Expansão da Malha de Transporte Dutoviário)
TIMES The Integrated MARKAL-EFOM system
UK United Kingdom
USA United States of America
toe tonnes of oil equivalent
VEC Vector Error Correction
YPF Yacimientos Petrolíferos Fiscales
YPFB Yacimientos Petrolíferos Fiscales Bolivianos
1
1 Introduction
1.1 The relevance of natural gas
Natural gas is considered as a transition energy carrier between high-carbon fossil
energy and renewable energy resources due to its low emissions, comparatively low-
capital cost conversion technologies and relatively abundant global reserves
(AGUILERA, 2014; IEA, 2011; RÍOS-MERCADO and BORRAZ-SÁNCHEZ, 2015).
In contrast to oil, natural gas reserve additions have consistently outpaced
production volumes and resource estimations have increased steadily since the 1970s
(IEA, 2011). The global natural gas resource base is vast and more widely dispersed
geographically than oil (BRUCKNER et al., 2014).
Since the early 1970s, world proven reserves of natural gas have been increasing
steadily, at an average annual rate of about 5%(ECONOMIDES and WOOD, 2009).
Considering the vast natural gas resources, according to ECONOMIDES and
WOOD (2009), it is conceivable that through further exploration success, changing
market conditions, investment in infrastructure and technological innovation, gas
reserves will increase over time, even though globally gas extraction and consumption
keep rising. Besides, according to ECONOMIDES and WOOD (2009), most
explorationist accept that it is easier to find more gas resources than oil resources,
therefore a scenario with gas resources scarcity is unlikely to occur at least for the next
50 years.
In 2015, 42.8% of total gas proved reserves (6599 Tcf) were located in the Middle
East, mainly in Iran (18.2%) and Qatar (13.1%). After the Middle East, Eurasia
concentrates the second highest proved reserves of natural gas, mainly in the Russian
Federation (17.3%) and Turmekistan (9.4%), followed by Asian countries (8.4%),
Africa (7.5%), North America (6.8%) and Latin America (4.1%) (BP, 2016).
2
Natural gas is also an important feedstock to produce other energy carriers such as
hydrogen (KATHE et al., 2016), electricity in power plants (VAN DEN BROEK et al.,
2015), medium distillates (SOUSA-AGUIAR et al., 2005) and even extra-heavy oils
(SENA; ROSA; SZKLO, 2013).
On the demand side, natural gas-fired power plants represented 22% of total
world electricity generation in 2012(EIA, 2016a) and the share of electricity production
from natural gas (NG) is expected to increase in the coming years due to the high plant
efficiency (relative to pulverised coal plants), low investment cost, high reliability and
flexibility (KHORSHIDI et al., 2016). Natural gas technologies play a key role for
supporting the security of supply in power systems with higher penetration of
renewables, as they are flexible and can react quickly to peak demands and other supply
sources shortages (DEVLIN et al., 2016).
In the transport sector, natural gas (primarily methane) can be compressed (CNG)
to replace gasoline in Otto-cycle (spark ignition) vehicle engines after minor
modifications to fuel and control systems. CNG can also be used to replace diesel in
compression ignition engines but significant modifications are needed(SIMS et al.,
2014). For heavy duty vehicles and ships, Liquefied natural gas (LNG) is also
considered to be a technological option (ARTECONI et al., 2010; EKANEM ATTAH;
BUCKNALL, 2015; KUMAR et al., 2011; STEFANA et al., 2016).
By its affordable price and low-emissions, natural gas is used extensively for
thermal energy needs in the industrial and buildings sectors (FRANCO, 2016) such as
for heating, drying and cooking. Furthermore, many empirical studies have found a
causal relationship on how natural gas consumption can boost economic growth
(FURUOKA, 2016; RAFINDADI and OZTURK, 2015).
Natural gas is also the fossil fuel that produces the lowest amount of GHG per
unit of energy consumed and is therefore favoured in mitigation strategies. Moreover, as
mentioned before, the role of gas-fired electricity generation as a complement to
variable renewables-based generation supports the prospect for growth in gas-fired
generation as a component of action to limit climate change(IEA, 2011). In addition,
3
natural gas has significant advantages compared to other fossil fuels and biomass in
terms of local pollutants such as CO, sulphur, particulates and unburnt hydrocarbons
emissions (SEDDON, 2006). NOx emissions are relatively high for natural gas
combustion because of the high flame temperature which assists the direct formation of
nitrogen oxides from the combustion air, however there are technological solutions to
reduce these emissions (ANDREWS, 2013).
Hence, given the recognised benefits of the natural gas use, many governments
have adopted policies to promote and support it, such as happened in Latin America,
specifically in the Southern Cone(CHAVEZ-RODRIGUEZ et al., 2016a; DICCO, 2013;
FAPESP, 2016; MINENERGIA, 2015; PLANALTO, 2000), which is the geographical
focus of this thesis.
1.2 Natural gas in the Southern Cone
The Southern Cone region encompasses the following countries: Argentina,
Bolivia, Brazil, Chile, Uruguay, and Paraguay. These countries have a varied energy
profile reflecting the size of the different countries, their population density, climate
conditions, availability of primary energy resources, economic structure and
development, degree of coverage of transport and distribution systems (FILIPPÍN et al.,
2012). The Southern Cone has a high level of participation in hydropower and non-
conventional renewables in the energy-mix and, for some countries (Chile and
Argentina), a pronounced seasonal behaviour of the natural gas demand.
Natural Gas has played an important role to diversify the energy mix in the
Southern Cone particularly in Argentina, Bolivia, Brazil, and Chile. Therefore, despite
its increasing importance in Uruguay, henceforth this thesis will focus on the following
countries: Argentina, Bolivia, Brazil, and Chile. This simplification is due to lack of
available data for Uruguay and Paraguay and also to their relatively small natural gas
consumption. In fact, the joint natural gas consumption of Argentina, Bolivia, Brazil,
4
and Chile represented 63% of the natural gas consumption in South America in
20141(BP, 2016).
As shown in Figure 1.1 natural gas has been the energy source with the higher
market share increase since 2000 in Argentina and Bolivia. From 2000 to 2014, natural
gas consumption in the Southern Cone has been increasing steadily at 4.6% per year,
mostly driven by the Brazilian market evolution that, from 2000 to 2014, has increased
almost fourfold (BP, 2016; MHE, 2015). There are significant differences in market
dimensions between the four SC countries studied in this thesis. In terms of natural gas
consumption, the major markets are Argentina and Brazil, consuming 1,848 TJ and
1,731 TJ, respectively in 2014 (EPE, 2015a; SECRETARIA DE ENERGIA, 2015a),
while Bolivia and Chile, have substantially smaller markets of 133 TJ and 159 TJ,
respectively in the same year (CNE, 2016a; MHE, 2015) (Figure 1.1).
Figure 1.1. Evolution of primary energy consumption in the four studied SC
countries. Source: CNE (2016a); EPE (2016a); MHE (2015); SECRETARIA DE
ENERGIA (2015a)
1 Other important consumers in the continent are Venezuela (20%) and Colombia (7%) (BP,
2016).
5
Figure 1.2. Natural Gas consumption structure by sectors in the Southern Cone in
2014. Source: CNE (2016a); EPE (2015a); MHE (2015); SECRETARIA DE ENERGIA
(2015a).
Although the structure of natural gas consumption is heterogeneous across these
countries, the industry and power generation sectors are the main natural gas consumers
in all of them (Figure 1.2).
Natural gas consumption in residential, commercial and public sectors in
Argentina and Chile have seasonal patterns driven by space heating during cold
seasons, as shown by CHAVEZ-RODRIGUEZ et al., (2016c). On the other hand,
industrial natural gas consumption is not clearly influenced by outdoor air temperature
and follows a flat pattern (SÁNCHEZ-ÚBEDA and BERZOSA, 2007). These annual
patterns are relevant in terms of energy systems modelling and natural gas models
should address them.
On the supply side, Figure 1.3 depicts the history of natural gas markets in the
Southern Cone. In the 1990s, there was a wave of market liberalization in South
America, which triggered investment to exploit natural gas resources, especially in
Bolivia and Argentina, not only in the upstream section, but also in the expansion of
domestic and international natural gas pipelines to trade with neighbouring countries.
0%
20%
40%
60%
80%
100%
Argentina Bolivia Brazil Chile
Residential Commercial and Public Industrial
Transport Power Generation
6
Figure 1.3. Natural gas net production2 and imports in the Southern Cone
countries. Source: Own elaboration based on ANP (2007, 2015); BP (2016); CNE
(2016); EIA (2015); INE (2014b); SECRETARIA DE ENERGIA (2015a).
During this decade, Brazil's natural gas industry expansion relied considerably on
the energy integration with South America, especially through natural gas (MATHIAS
and SZKLO, 2007). The construction of 3,150 km of pipelines connecting Bolivia’s gas
sources with the south-eastern and southern regions of Brazil started in 1997 and began
operating in 1999. The so called GasBol pipeline reached its full capacity in 2005
(30 million m3/day).
Furthermore, in the south, important investments in transport infrastructure were
made, allowing connecting Chilean consumers at the north, centre and south of the
country with the productive Argentinean basins during the end of the 1990s. As a result,
seven pipelines with a combined length of over 3,500 km were built.
Then, in the 2000s, due to the lack of investment in the upstream sectors of
natural gas and the high demand in Argentina, the Kirchner's administration decided to
suspend the exports of natural gas in order to keep the internal demand satisfied. The
2 Gross natural gas production discounting flare, reinjection and self-consumption.
7
gas crisis peaked in 2007 when Argentinean exports were zero. Figure 1.3 illustrates
how natural gas net production and imports have evolved in the Southern Cone
countries.
To secure natural gas supply, Chile constructed two LNG regasification terminals
with a 5.7 MTPA (metric tonnes per annum) total regasification capacity (IGU, 2015).
In the same direction, to ensure supply through diversification, Brazil and Argentina
built LNG regasification plants, with a total capacity of 11.7 MTPA and 7.6 MTPA
respectively (IGU, 2015). As a result, LNG imports increased their market share in
Argentina, Chile and Brazil. Figure 1.4shows the geographic locations of the natural gas
pipelines infrastructure and LNG regasification plants.
Figure 1.4. Natural Gas Transport and Regasification Infrastructure in the
Southern Cone.
8
There are many uncertainties about the future of natural gas market in the
Southern Cone region, both in the supply and demand sides. These uncertainties are
driven by the role that the following key factors will play in the natural gas market:
- The unconventional gas resources in Argentina.
- The associated gas from offshore oil fields in Brazil.
- The penetration of LNG.
- The international trade of natural gas among Southern Cone countries.
- The increasing share of renewables technologies in the power generation
mix.
Assessing these key factors is pivotal for future energy policy-making in the
Southern Cone.
1.3 Research Question, Aims and Objectives
The overarching objective of this study is to model the natural gas markets in the
Southern Cone to provide quantitative insights about the future development of these
markets and the key factors influencing them. In order to address this objective, it is
worth answering more specific research key questions:
- What is the economic potential of unconventional resources in
Argentina?
- Will Brazil be gas self-sufficient using its associated gas production from
offshore fields?
- What role LNG will play in the natural gas market of the Southern Cone?
- Will natural gas trade in the Southern Cone increase with the current
pipeline infrastructure?
- Will renewable energy affect the natural gas consumption in power
generation?
As important as exploring these issues are the construction of a multi-country
natural gas model underpinned by economic and technology fundamentals of the gas
industry. This study does not address the national regulation structure for each country,
9
neither macroeconomic structures nor geopolitical aspects but is focused on the
fundamentals of supply and demand of natural gas. Finally, it is relevant to say that this
thesis pretends to fill a gap in terms of scientific literature in other-than-Brazil Latin
American countries.
1.4 Overview
The remainder of this thesis is structured as follows: Chapter 2 provides a
literature review on natural gas modelling using energy planning tools and the key
factors that can shape the future natural gas market. The research design and the energy
planning tools used for this thesis are depicted in Chapter 3. Chapter 4 focuses on the
natural gas demand projections for end-uses. In Chapter 5 the power generation sector
and the natural gas chain in the upstream and midstream modelling is explained. This
chapter also describes the assumptions made for the scenario construction. Chapter 6
shows the projections of natural gas domestic production, the natural gas supply and
demand for the two scenarios modelled. Chapter 7 tries to answer the thesis’ research
questions detailed before, by using sensitivity scenarios to provide additional
quantitative insights. As such, it analyses the role of each key factors influencing the
future natural gas markets in the Southern Cone. Finally, Chapter 8 provides the final
remarks and the limitations of the study, also highlighting the opportunities for policy-
making.
10
2 Literature Review
2.1 Natural gas models
Several global, regional and country-level studies have looked at the medium- and
long-term role of natural gas (from primary energy production to conversion into energy
carriers as electricity and end-use consumption). Table 2.1shows the main features of
different energy planning natural gas models found in the literature. The energy
planning natural gas models assessed, which involves the modelling of the whole
natural gas chain and sometimes as part of an integrated energy planning tool, differs
from specific engineering gas models. For instance the energy planning tools do not
involve fluid mechanics fundamentals (e.g., loss of pressure, compressors required, etc.)
but on the other hand provides results from the upstream to the demand of natural gas
and its interplay with other energy supply sources.
There are commercial natural gas models extensively used in the industry such as
Wood Mackenzieʼs Global Gas Model (GGM), Deloitte ʼs Market Point World Gas
Model, Nexant World Gas Model, etc. However, these models are not assessed in this
section as there is little public information about them and they are out of scope of this
academic study.
11
Table 2.1. Review of natural gas modeling in Energy Planning
Model Name Approach
(Top-down
/Bottom-
up)
Simulation/O
ptimization
Geographi
cal Scope
Geographical
Resolution
Temporal
Horizon
Temporal
Resolution
Does it
model
other
sectors?
Objective Comments Sources
The Rice
World Gas
Trade Model
Top-Down Optimization World By country
(larger countries
subdivided by
regions)
2011-2040 Yearly No Maximize the present
value of producers rents
within a competitive
framework
Develops reserves from existing fields and undiscovered deposits,
constructs pipelines and LNG delivery infrastructure, and
calculates prices to equate demands and supplies
(HARTLEY;
KENNETH,
2005)
The World
Gas Model
Top-Down Optimization
(mixed
complementar
ity
formulation)
World By country
(larger countries
subdivided by
regions)
2005-2040 5 years
(seasons
customized
)
No Maximize the
discounted income
calculated as the
difference between
revenues and
production costs of
each producer
separately
Allows to take into account endogenous investment decisions in
infrastructure over the next decades while at the same time
including market power in the pipeline and the LNG market
(EGGING;
HOLZ;
GABRIEL,
2010)
ETP-TIMES Bottom-up Optimization World Aggregated in
28 regions
2010-2075 5 years (a
year
divided 4
seasons)
Yes Minimization of the
total system costs
The model covers from primary energy supply and conversion to
final energy demand of all sectors. Technologies are described by
their technical and economic parameters, such as conversion
efficiencies or specific investment costs. Learning curves are used
for new technologies to link future cost developments with
cumulative capacity deployment.
(IEA, 2016a)
ETSAP-
TIAM
Bottom-up Optimization World Aggregated in
15 regions
2000-2100 5 years Yes Minimization of the
total system costs
The model contains explicitly detailed descriptions of more than
one thousand technologies and one hundred commodities in each
region, the chain of processes which transform, transport,
distribute and convert energy into services from primary resources
in place to the energy services demanded by end-users
(IEA, 2016b)
EPPA Top-down Optimization World 16 regions 2000-2100 5 years Yes General Equilibrium Simulates the world economy through time to produce scenarios of
12
(Computable
General
Equilibrium
model)
(includes
other
economic
sectors)
greenhouse gases, aerosols, other air pollutants and their
precursors, emitted by human activities.
PET Bottom-up Optimization EU27+ 9
other
European
countries
By Country 2005-2050 5 years (4
seasons and
a 3 intra-
day periods
Night, Day
and Peak)
Yes Minimization of the
total system costs
Encompasses all the steps from primary resources in place to the
supply of the energy services. Energy policies such as investment
subsidies, feed-in tariffs, renewable quota systems (biofuel, other
renewable energy), all according to the specific details for each of
the Member States were modeled.
(KANORS,
2016)
PRIMES Bottom-up Optimization Europe (35
European
countries)
By country 2010 -2050
By
5- years
steps
Daily basis,
a few
typical days
are
rep
resented
per country
Yes, it
covers all
the energy
system
Agents act individually
optimizing their profit
or welfare using
individual(private)
discount rates for
capital-budgeting
choices
Provides detailed projections of energy demand, supply, prices and
investment to the future, covering the entire energy system.
Detailed representation of gas infrastructure (field production
facilities, pipelines, LNG Terminals, Gas Storage, Liquefaction
Plants).
(E3MLAB/ICC
S, 2014)
GASTALE Top-down Partial
Equilibrium
Europe By country Yearly Yearly No Partial Equilibrium Describes gas market in terms of two layers of companies on the
supply side along with consumers in three basic sectors on
the demand side of the market. The market structure is
assumed to consist of an oligopoly of upstream gas producers and
a layer of downstream gas traders, all of whom seeks to maximize
its profits , the position of traders can vary from a national
monopoly to perfect competition between traders.
(BOOTS;
RIJKERS;
HOBBS,
2004)
GASMOD Top-down Partial
Equilibrium
Europe and
exporters(R
ussia,
Algeria,
Middle
East,
By country Yearly
(2003 base-
year)
Yearly Ño Partial Equilibrium A static model which structures the natural gas market as a two-
stage game of successive i) exports to Europe, and ii) trade within
Europe. In contrast to GASTALE, it incorporates an endogenous
determination of domestic production. Infrastructure capacities are
explicitly taken into account in the model.
(HOLZ; VON
HIRSCHHA
USEN;
KEMFERT,
2008)
13
Nigeria)
TIGER Bottom-up Optimization Europe By node (500
nodes across
Europe and 650
pipeline
sections)
Ten years Monthly
resolution
No Least operation cost Dispatch model optimizing natural gas supply to all European
countries subject to the available infrastructure. Total costs
comprise of transport, storage and production costs It assumes
perfect competition.
Strategic
Eurasian
natural gas
market
Top-Down Partial
Equilibrium
(mixed
complementar
ity problem)
Eurasia By country 2010-2030 Yearly No Satisfy each market
participant’s Karush-
Kuhn-Tucker
conditions for profit
maximization and
market clearing
conditions
The model represents horizontal oligopolistic relationships among
producers, bilateral market power between producer (Russia) and
transit (Ukraine) countries, detailed transport constraints, and
operation decisions.
(CHYONG;
HOBBS,
2014)
NEMS Bottom-up Partial
Equilibrium
(Heuristic
algorithm to
establish a
market
equilibrium)
United
States
12 U.S and 2
Canadian
demand region
(NGTDM)
12 U.S.
Supply Region
(OGSM)
2014(or
year-base)-
2040
Two
seasons of
each
forecast
year.
Yes, it
covers all
the energy
system
(NEMS)
Solves prices that will
balance the quantities
producers are willing to
supply with the
quantities consumers
wish to consume.
Gas-Associated production linked to oil prices and fixed to oil
production. Non-Associated natural gas production driven by price
to provide equilibrium between supply and demand. Flows of
natural gas and expansion through a regional interstate
representative pipeline network, for both a peak and off-peak
period. Allow regulation constraints and input taxes (royalty, state
and federal taxes, etc.)
(EIA, 2013a)
(EIA, 2014)
UKTM Bottom-up Optimization United
Kingdom
UK single
region
2010-2100 5 years (4
seasons and
4 intra-day
periods:
night, day,
evening
peak, late
evening)
Yes Minimization of the
total system costs
Technology-oriented, dynamic, linear programming optimisation
model representing the entire UK energy system from imports and
domestic production of fuel resources, through fuel processing and
supply, explicit representation of infrastructures, conversion to
secondary energy carriers, end use technologies and energy service
demands.
(DALY;
FAIS, 2014)
14
The models presented in the Table 2.1 have been applied for different purposes,
either considering natural gas as the mains focus of the research or including it as a part
of the whole energy system.
At a global level, the Rice World Gas Trade Model was early applied to assess
Russia’s ability to influence the world natural gas market (HARTLEY and MEDLOCK,
2009). Also, supporting a study of the U.S. Department of Energy, this model was
applied to investigate how the development of extensive global shale gas resources
could alter geopolitical relationships over the coming decades and map out implications
for U.S. energy security (MEDLOCK et al., 2011).
The World Gas Model of University of Maryland has been used to analyse
possible future gas cartels and their effects on gas markets in a number of regions across
the world (GABRIEL et al., 2012). Furthermore, HOLZ et al. (2015) used this model to
explore the key drivers and uncertainties for natural gas markets in the coming decades
such as the production of conventional and unconventional sources, the role of
international trade, and the impact of climate policies. Lately, a stochastic version of the
model (EGGING, 2013) has been used to investigate the impact of several possible
risks on the European, North American, and Asian gas markets.
The IEA (IEA, 2015a) has been using the ETP-TIMES optimization energy
system model to develop scenarios up to 2050. It concluded that, from the perspective
of reducing greenhouse gases emissions (GHG), natural gas power plants are a good
replacement for coal power plants up to 2030, and the increased use of natural gas in
shipping and aviation will allow further (and necessary) emission abatement.
GRACCEVA and ZENIEWSKI (2013) used the global energy system model
ETSAP-TIAM to analyse how shale gas can impact regional gas production, inter-
regional trade, demand, and price until 2050. The authors have found that in “an
optimistic but carbon-constrained global energy system” the room for increased
importance of shale gas production would be lower vis-à-vis a non-carbon constrained
case. Nonetheless, an increase in the relative importance of natural gas “confirming a
role for gas as a cost-effective bridge towards a low-carbon energy system” would
15
occur. The authors estimated natural gas supply curves for the USA, China, and
Western Europe and computed regional natural gas consumption including the Southern
America.
PALTSEV (2014) applied the general equilibrium economic model EPPA to
explore scenarios of Russian natural gas exports up to 2050. This author found that EU
still will be important natural gas trading partners for Russia in the decades to come and
exports will increase to the Asian markets. However, shale gas developments in China
and Europe and LNG supplies might undermine Russian exports.
The gas market system for trade analysis in a liberalising Europe (GASTEL) was
used to assess the role of the downstream trading companies and their interactions with
oligopolistic producers in Europe (BOOTS et al., 2004). By applying a similar model
for Europe, GASMOD, (HOLZ et al., 2008) found that Cournot competition is the most
suitable representation of the European natural gas market.
A geographically disaggregated bottom-up model, TIGER (Transport
Infrastructure for Gas with Enhanced Resolution), was used to assess the impacts of the
Russian-German Nord Stream(Baltic Sea) import pipeline on the Europe’s gas system,
especially volume flows within the grid and the utilization of import pipelines
(LOCHNER and BOTHE, 2007). Results suggest that the Nord Stream pipeline mainly
jeopardizes the transport volumes in the traditional transit pipelines transporting
Russian Gas. CHYONG and HOBBS (2014) also assessed the Nord Stream pipeline,
using the Strategic Eurasian Natural Gas Model, and found that this project is profitable
for investors and improves social welfare in all market power scenarios. They also
found that the project is sensitive to the degree of downstream competition in European
markets. The lack of downstream competition might result in the negative value of the
Nord Stream system to Gazprom.
Based on a top-down approach from a global to a regional level, the research
project REACCESS (2011) used two energy system optimization models, TIAM (for
the world) and PET- Pan European TIMES (for EU27+), to evaluate technical,
economic and environmental features of existing and future energy corridors within
Europe and between European countries and the rest of the world. A detailed
16
representation of the natural gas supply chain was modelled in PET including re-
gasification plants to take into account the liquefied natural gas imported by ships and
seasonal natural gas trade exchanges (REACCESS, 2011).
Other authors such as CAPROS et al. (2012a, 2012b) used the PRIMES (Price-
Induced Market Equilibrium System) partial equilibrium energy system model to assess
the decarbonisation of the EU energy system until 2050. They concluded that EU power
sector can reduce its CO2 emissions by 98% with respect to 1990 levels, by replacing
coal and gas power plants with renewable electricity and carbon capture and storage
(CCS) gas plants. In a more recent study, CAPROS et al. (2014) performed a multi-
model analysis (PRIMES, GEM-E3, and TIMES Pan EU) to assess the reduction of EU
GHG emissions by 80% in 2050 compared to 1990 levels. In this case, the largest part
of the emission reduction effort until 2050 relies on the substantial deployment of CCS
technologies after 2030, as well as natural gas replacing coal and oil in the primary
energy mix. Up to 2030, natural gas power plays a major role in EU, as cost-effective
CCS is not foreseen to be available soon.
At the national level, the National Energy Modelling System (NEMS) has been
used by the Energy Information Administration (EIA) to model energy markets in the
United States of Americas. NEMS is a benchmark model for Partial Equilibrium
Integrated Energy Planning. For the natural gas chain, NEMS contains an Oil and Gas
supply module, which has 12 production regions (EIA, 2014). The regional forecasts of
natural gas wellhead prices and production are provided by the Natural Gas
Transmission and Distribution Module (NGTDM), which encompasses twelve U.S
regions and 2 Canadian regions(EIA, 2013a). NEMS is the main tool for the elaboration
of the Annual Energy Outlook.
In addition, another model is the UKTM (UK Times model) of the University
College London has been extensively used in the United Kingdom. This model was
used in MCGLADE et al., (2016) to assess how much gas use is compatible with
meeting the UK’s carbon emissions reductions targets. The authors concluded that gas
is unlikely to act as a cost-effective ‘bridge’ to a decarbonised UK energy system. The
analysis shows that gas could only act as a bridge from 2015-20, and CCS is required to
increase the use of gas in the long-term.
17
In general, modelling the natural gas chain inside Integrated Energy Models and
the construction of models specific to natural gas have advanced in the past decade.
From this literature review, it is possible to say that natural gas modelling has been used
for different purposes: assessing natural gas geopolitics (BOOTS et al., 2004;
CHYONG and HOBBS, 2014; GABRIEL et al., 2012; HARTLEY and MEDLOCK,
2009); understanding natural gas role under the climate change mitigation (CAPROS et
al., 2012a, 2012b, 2014; IEA, 2015a; MCGLADE et al., 2016); analysing key drivers of
natural gas markets (EGGING, 2013; GRACCEVA and ZENIEWSKI, 2013; HOLZ et
al., 2015; LOCHNER and BOTHE, 2007; MEDLOCK et al., 2011; PALTSEV, 2014;
REACCESS, 2011); and simply forecasting natural gas supply and demand balance
(EIA, 2013a, 2014).
In addition, some models assume a perfect competition market (DALY and FAIS,
2014; E3MLAB/ICCS, 2014; EIA, 2014; HARTLEY and KENNETH, 2005; IEA,
2016a, 2016b; KANORS, 2016), while others simulate imperfect competition, such as
oligopolistic behaviours (BOOTS et al., 2004; CHYONG and HOBBS, 2014; EGGING
et al., 2010; HOLZ; VON HIRSCHHAUSEN and KEMFERT, 2008).
Finally, it is relevant to note that several natural gas studies at the regional level
have focused on European markets and its interplays with Russian exports (BOOTS et
al., 2004; CHYONG and HOBBS, 2014; HOLZ et al., 2015; LOCHNER and BOTHE,
2007; REACCESS, 2011). There is no study that focused on the Southern Cone.
2.2 Southern Cone Gas Market Studies
This section aims at reviewing the literature that focuses on the Southern Cone
gas market integration.
JADRESIC (2000) detailed the benefits of a pipeline over the Andes from
Argentina to Chilean terms of lower energy prices, higher environmental standards, and
a more reliable energy system. The study also emphasized how technological
18
innovation, economic deregulation, and regional integration allows for building the
major international gas pipeline project within a competitive framework and without
direct State involvement.
KOZULJ (2004) depicted the regulatory and market framework for natural gas
integration initiatives suggesting that in the mid-term LNG would be an interesting
alternative for integration. Then, after the gas crisis in Argentina, KOZULJ (2008)
described how the new LNG projects in the region were not diverted towards South
American integration but responded to globalization strategies of energy security.
MARES (2006) analysed three historical cases associated with the natural gas
integration in the Southern Cone: Yabog, GasAnde sand GasBol pipelines. He aimed at
understanding why there was a twenty-five-year lag between the first international
pipeline project and the others, and to uncover the key factors that determine why some
pipeline projects were built while similar pipelines proposed languished. In another
study, MARES and MARTIN (2012) analysed the Chilean-Argentina integration, with
the argument that neither markets nor political will of leaders can produce successful
economic integration unless the politics of integration have been favourably resolved.
The GasBol project also attracted scientific research, especially from Brazilian
institutions. SANTOS et al. (2004) described the integration and impacts in Brazil of the
Bolivia gas supply. Analysing the Brazilian-Bolivian markets dynamics, HALLACK
(2007) explains how these markets are not based on gas-price dynamics such as the spot
market prices but on long-term contracts, the structure of assets owning, and the direct
actions for conflicts resolutions from the governments. GLACHANT and HALLACK
(2009) analysed the Brazilian-Bolivian case to underpin a contract life-cycle approach
to assessing the robustness of long-term take-or-pay contracts. In an ex-post analysis,
FUSER (2011) highlighted the foreign policy crisis that the Bolivian gas nationalization
brought to Brazil and how this damaged other Latin American integration initiatives.
Finally, RUDNICK et al. (2014) concluded that opportunities for deeper regional
integration in the Southern Cone are plentiful, although exploiting them will require
surmounting multiple institutional barriers, both nationally and internationally. In
addition, the authors highlighted the seasonality of Latin American natural gas demand,
19
which does not follow the international natural gas demand patterns, and how this can
compose an opportunity to integrate the region into the worldwide LNG network.
In terms of regulation, ALMEIDA and OLIVEIRA (2000) argued that Brazil has
not adopted the franchised monopoly of North America and Europe to develop the gas
supply industry in its infancy. MATHIAS and SZKLO (2007) assessed the reforms in
the Brazilian natural gas industry highlighting the necessity of a new legal framework to
encourage investments in networks and guarantee supply. COLOMER and HALLACK
(2012) analysed the new law in Brazil for the natural gas sector, passed in March 2009
(Law No. 11,909), suggesting that it is not sufficient to guarantee stimulus to the entry
of new players in the transportation segment. LION (2015) analysed the Brazilian
economic and regulatory framework for the development of unconventional gas
resources. For this purpose, LION (2015) used a discount cash flow model to conclude
the feasibility of the unconventional gas production in Brazil and to suggest regulatory
measures promoting the successful exploitation of these resources.
In terms of energy planning and modelling, KELMAN (2009) presented the
mathematical modelling for the integrated planning operation and expansion between
natural gas and power generation sectors.
The work of CUNHA (2010) is one of the first initiatives that studied possibilities
of gas integration in the Southern Cone and the effects of LNG in the natural gas
markets of this region. The author built a top-down model using a Cournot-Enke
formulation with a remarkable spatial disaggregation: 34 markets, connected by 77
possible trade routes through pipeline’s and 47 through LNG. However, uniform costs
in the supply just differentiated by onshore and offshore categories, and a total inelastic
demand undermined the capacity of the model to represent the economics and
engineering fundamentals of the natural gas markets.
DOS SANTOS et al. (2011) developed a set of tools to perform thermohydraulic
simulation of gas flow through pipelines, a Monte Carlo simulation for compressor
station availability studies, an economic risk evaluation related to potential revenue
losses and contractual penalties and linear programming for the maximisation of income
and the minimisation of contractual penalties.
20
DI SBROIAVACCA )2013) assessed unconventional resource in Argentina and
build scenarios for supply and demand of natural gas in Argentina based on different
levels of shale gas production. Using these Argentinian scenarios, CHAVEZ-
RODRIGUEZ et al. (2016a) assessed how Bolivian natural gas supply could behave,
particularly foreseeing if Bolivia would attend its export commitments.
Other interesting studies in the context of the Southern Cone region include the
work of BRAGA (2014), who proposed alternative methodologies to project the
industrial natural gas demand in Minas Gerais, Brazil; SANTOS (2015), who analysed
alternatives to monetize onshore natural gas resources in Brazil; ROCHEDO et al.
(2016), who explored capture and storage alternatives for the CO2 embedded in the
Brazilian associated gas from Pre-Salt oil production; and DIUANA et al. (2016) who
assessed the convenience of LNG long-term contracts vs the currently imports in Brazil
based on spot markets.
Brazil’s Ten-Year Expansion Plan for the Gas Pipeline Transport Network
(PEMAT) of EPE (EPE, 2014) is an emblematic study, in terms of natural gas
infrastructure planning at national level. In this study, the feasibility of different
proposed pipelines is analysed, according to natural gas supply and demand scenarios,
environmental and hydraulic assessments and economic analysis of each pipeline.
Finally, CAF/CIER (2012) identified interconnection opportunities in the
Southern Cone. The South American Model Base (SAMBA), built in OSeMOSYS, was
used to model electricity supply and transmission links between South American
countries MOURA and HOWELLS (2015), however the work of POSTIC (2015) who
modelled the South America energy market using TIMES was the only study identified
in this thesis that included the modelling of natural gas at a regional level involving the
Southern Cone, at a reasonable geographical resolution. Despite the big efforts made by
POSTIC (2015), the modelling approach was aggregated and simplified since this was
not his focus , but to assess national climate policies.
Therefore, there is a literature gap in natural gas market modelling and planning
studies at the regional level for the Southern Cone. Studies of this kind might contribute
21
to understanding the features of the natural gas market in the region and be a first-step
for further natural gas integration.
2.3 Key factors for the Southern Cone natural gas market
This section provides a historical framework for each of the key factors that might
drive changes in the Southern Cone natural gas market. These key factors were listed in
the previous chapter and will be discussed in chapter 7, after the runs of the simulation
model.
2.3.1 Argentina’s hydrocarbon industry and its unconventional resources
Oil production in Argentina started in 1907 in Comodoro Rivadavia. In 1922, the
State Oil Company, Yacimientos Petroleros Fiscales (YPF), was created. By the end of
the 1940s, YPF had monopolistic control of oil production. However, the State oil
company had neither the financial nor the technical capacity to develop the oil resources
in the country (GADANO, 1998). During the 1950s, the government fostered the
incorporation of private capital in the oil industry through exploitation and drilling
contracts. These measures had a positive impact on the oil production, mostly focused
on the Golfo San Jorge Basin (ECONLINK, 2008). The country became self-sufficient
at the beginning of the 1960s, but in the following years, the oil industry fluctuated
between nationalism and openness to private capital regimes (GADANO, 1998).
Natural gas started production commercially in the late 1940swhen the first natural gas
pipeline connecting Comodoro Rivadavia fields to Buenos Aires was constructed.
In 1977 the mega gas field of Loma de la Lata was discovered (14 TCF) in
Neuquen. This changed the country’s energy mix and major natural gas transport
pipelines were constructed in the 1970s and 1980s (RISUELO, 2010). Despite the
significant development in the Neuquen Basin, oil production started to decline, mainly
because of oil and oil products domestic price control and poor macroeconomic
conditions(ECONLINK, 2008; GADANO, 1998). This motivated in the late 1980s a
market-oriented reform in the oil and gas industry, during Menem administration. The
22
constraints on foreign trade were eliminated and private companies became able to
freely manage the returns on their oil or gas investments (VÁSQUEZ, 2016).
Furthermore, in 1993, YPF was privatized. The country raised its oil production, which
started to decline again in1995. In 2012, the Argentine congress passed the
Expropriation Law. Among other matters, the Expropriation Law provided for the
expropriation of 51% of the share capital of YPF from Repsol (YPF, 2016).
In the 1960s and 1970s, YPF discovered unconventional resources in the Puesto
Hernández and Loma La Lata fields and drilled the Vaca Muerta and Los Molles
formations. However, neither prices nor technology allowed the production. However,
according to DI SBROIAVACCA (2013), the first shale well drilled in Loma La Lata
by YPF in 2010 can be considered as the landmark of the unconventional resources in
Argentina.
Unconventional gas wells are a set of different formations, categorised as tight
sands, shale gas and coal bed methane. Tight sand and shale gas reservoirs differ from
conventional gas as they exhibit a low permeability and low porosity, which makes it
harder to extract the gas. Stimulation is required to produce a sufficient gas flow, such
as hydraulic fracturing (and some acidizing methods). Unlike tight sands, horizontal
drilling is also normally needed for shale gas, in order to get access to more of the gas
(SGI, 2015).
Argentina has both tight gas and shale gas reservoirs. In 2015, only 4 areas of the
16 producing areas overcame the output of 500 Mm3/year: Loma La Lata-Sierra
Barrosa, Rincon del Mangrullo, Lindero Atravesado and El Mangrullo. In terms of shale
gas, Loma Campana field represented 59% of total shale gas production in 2015,
followed by El Orejano and Aguada Pichana, with 21% and 14%, respectively. In
December of 2015, the tight gas production in Argentina summed 445 Mm3/month,
with 432 producer wells, whereas the shale gas production summed 108 Mm3/month,
with 527 producer wells (IAPG, 2016). It is expected that unconventional gas
production will increase over time, as Argentinian unconventional natural gas resources
are evaluated as the second larger accumulation in the world after China (EIA, 2013b).
23
Therefore, the degree of success for tapping unconventional resources in
Argentina can affect the natural gas dynamics in the Southern Cone. This is a major
influence key factor, then.
Figure 2.1. Shale resources in Argentina.
Source: GARRISON (2016)
Therefore, a major question to be addressed is: how the development of
unconventional gas in Argentina can impact the dynamics of the natural gas markets in
the Southern Cone? This question is addressed in Chapter 6 where the main results for
the Southern Cone are presented. A zoom analysis on unconventional gas is performed
in chapter 7 in an attempt to answer the question about what is the economic potential
of unconventional natural gas resources in Argentina.
2.3.2 Petroleum production in Brazil, the offshore associated gas, and the Amazonian
resources
Brazil started its petroleum production in the 1940s in the Recôncavo Basin. In
1953, the Brazilian State Oil Company, Petrobras, was created to monopolize the oil
activities in upstream, midstream and downstream. During the1950s, upstream activities
24
in the Amazon were intensified and in the Sergipe-Alagoas Basin, oil accumulations
were discovered. The onshore production in the Sergipe-Alagoas Basin started in the
1960s and all the onshore basins were explored (LUCCHESI, 1998).
In the 1970s, production started in the Espirito Santo and the Ceará-Potiguar
Basins. Nevertheless, the discoveries of Garoupa and Echova fields in the Campos
Basin in 1974 and 1977, respectively, were the milestone for the beginning of offshore
activities in Brazil (DA COSTA et al., 2003). The giant deep-water fields Albacora and
Marlim were discovered in 1984 and 1985, located between 300 to over
1000m.depthAlbacora started production in 1987 and Marlim in 1990 (DA COSTA et
al., 2003). According to SZKLO et al., 2007), technology progress, such as
development of 3-D and 4-D seismic; development of directional and horizontal
drilling; and floating rigs, especially floating production, storage and offloading(FPSO),
etc. were the main driving forces behind the increase in reserves and production in
Brazil. Other relevant post-salt discoveries include Albacora Leste (1986), Marlim Sul
(1987), Marlim Leste (1987), Barracuda (1989), Caratinga (1989), Espadarte (1994),
Roncador (1996) and Cachalote (2002) (DA COSTA et al., 2003).
In 2006, Petrobras was successful in drilling and discovering oil through a
massive salt layer off the Brazilian coast that stretches from the Campos to the Santos
Basin, thereafter known as pre-salt reservoirs. The pre-salt province occupies an area of
approximately 149,000 km² and the reservoirs are located in ultra-deep waters at total
depths of up to 7,000 meters. In the Santos Basin, the salt layer is approximately two
kilometres thick. In the northern part of the pre-salt province, the salt is thinner and
much of the oil has migrated through the salt to the post-salt sandstone reservoirs of the
Campos Basin. Pre-salt production began in 2008 in Campos Basin from the Jubarte
field, and in 2010 in the Santos Basin from the Lula field (PETROBRAS, 2016) .
Still, post-salt production in Campos Basin is currently the major source of oil in
Brazil. Roncador field was the largest post-salt oil producer, with an average of 300
kbbl/day in January of 2016, followed by Marlim and Marlim Sul with 157 kbbl/day
and 146 kbbl/day, respectively. Nevertheless, pre-salt fields production has been
increasing at a fast pace. Currently, Lula is the top oil field producer in Brazil, with an
average of 405 kbbl/day in January of 2016, and other two pre-salt oil fields, Sapinhoá
25
(206 kbbl/day) and Jubarte (173 kbbl/day), are the top 3 and top 4 oil producer in the
country. In January of 2016, there were ten pre-salt producing oil fields(MME, 2016a).
In terms of natural gas production, in January of 2016, Lula was the field with the
highest associated natural gas production (Figure 2.2). Nonetheless, post-salt production
remains the major source of natural gas in Brazil. In 2015, pre-salt fields produced 17.1
Mm3/day whereas post-salt offshore fields produced 56.15 Mm3/day (ANP, 2016a;
MME, 2016a). Most of this natural gas comes as associated to oil production and, in the
case of offshore fields, it is connected to natural gas processing plants in the coast
trough pipelines. Some pre-salt fields have a high gas to oil ratio, such as Lula and
Sapinhoá, with 0.021 Mm3/Mbbl and 0.029 Mm3/Mbbl, respectively, based on their
production of January (MME, 2016a).
Figure 2.2. Top natural gas production fields in Brazil in January, 2016. Source:
(MME, 2016b)
To increase natural gas production in Brazil, Petrobras has focused on pre-salt
fields., The company is investing in transferring pipelines from offshore facilities to the
coast, is building a new natural gas processing plant in the Comperj complex, and is
expanding the capacities of existing natural gas processing plants (NGPP) near to the
fields (PETROBRAS, 2016, p. 20). Moreover, the CO2 content in the associated natural
0
5
10
15
20
25
Lula Mexilhão Sapinhoá Leste doUrucu
RoncadorRio Urucu Manati GaviãoReal
Jubarte MarlimSul
MMm3/day
Associated gas Non-Associated gas
26
gas from pre-salt fields is estimated in a range from10% to 45% mol, which makes it
unfeasible to transport the gas to the coast (SILVA, 2015). Consequently, additional
investments are required to separate and inject the CO2 removed into the reservoirs.
On the other hand, there are relevant offshore non-associated natural gas fields
such as Mexilhão in the Santos Basin and Manati in the Camamu Basin (Bahia), both of
them operated by Petrobras, and onshore fields such as Gavião Real operated by
Parnaíba Gas, a private gas company.
Brazil has relevant unconventional gas resources. According to EIA (2013b),
there are technically recoverable unconventional gas resources estimated in 245 Tcf in
the country. However, difficulties in mobilizing resources for unconventional gas
production due to the shortage of investments and capacity of industrial services, along
with water availability restriction for hydraulic fracturing in some basins, might limit
the technical potential of the activity (CAMARGO et al., 2014).
In the North of Brazil, current hydrocarbon production is focused on the Urucu
area, a highly-biodiverse area within the Solimões Basin, in the Amazonas State, and
the largest onshore natural gas reserve in Brazil. The commercial discovery of oil and
gas in this area dates from 1986. In 1988 the commercial production of light oil started
(PETROBRAS, 2010). Oil production is pumped to the Coari Terminal and then
transported by vessels to the Refinery Isaac Sabá (REMAN) in Manaus. LPG obtained
from associated natural gas liquids follows the same route to the Coari Terminal, and
then supplies cities of Amazonas, Pará, Rondônia and Maranhão.
Initially, due to the lack of a natural gas transport infrastructure and market, the
associated natural gas was processed at the NGPP of “Polo Arara” to obtain the natural
gas liquids, and the dry-gas was re-injected into the fields. In 2009, the natural gas
pipeline Urucu-Coari-Manaus started operation with a transport capacity of 5.5
Mm3/day. This pipeline connects “Polo Arara” in Urucu to Manaus trough a route of
663 km; besides, there are 139 km of pipelines in 9 branches to Coari (PETROBRAS,
2015a). The natural gas transported to Manaus is mainly used to fire the thermal power
plants owned by Electrobras. It is also used by some industrial facilities. According to
27
Cigás, the natural gas distribution concessionary, these facilities consumed only 3.5
Mm3/day in 2015 (CIGÁS, 2015).
According to EPE (2015b), new natural gas power plants are planned for the
Northern Region of Brazil and it is also expected an increase in natural gas production
from the Solimões Basin. Natural gas production in Solimões Basin is mainly associated
with the oil production fields, however, non-associated natural gas production has
increased in the last years (ANP, 2015). There are estimates of technically recoverable
shale gas resources in Amazonas and Solimões Basin of 100 and 65 Tcf (EIA, 2013b).
However, tapping these unconventional resources is not considered in the time-horizon
of the study.
Despite the significant natural gas resources in the Brazilian Amazon, as seen in
Figure 2.2, associated gas from offshore oil fields are the major sources of domestic
natural gas production in Brazil. This gas production is expected to increase as Pre-Salt
oil fields are presenting high yields, and much of the gas produced is being reinjected
not only by processing infrastructure constraints but also because of lack of market.
Testing the limits of the potential of these domestic resources, “will Brazil be gas self-
sufficient using its associated gas production from offshore fields?”.
2.3.3 Regional trade of natural gas: the Argentina-Chile case and the Bolivian exports
The privatization process of the natural gas industry in Argentina in the1990s
carried out important investments in transportation infrastructure, allowing the
interconnection of Chilean consumers at the north, centre, and south of the country to
the productive Argentinean basins.
As a result, over 3,500 km of pipelines were built, with an investment of over 2
billion dollars. Seven international pipelines between Argentina and Chile were built in
the 1990s: Bandurria (2 Mm3/d), Dungeness (2 Mm3/d) and El Condor-Posesión (2
Mm3/d) in the South, Gasandes (9.5Mm3/d) and Pacifico (3.5 Mm3/d) in the centre,
Atacama (9 Mm3/d) and Norandino (5 Mm3/d) in the North.
28
Due to the economic crisis in Argentina, natural gas tariffs were frozen in October
2001, as a measure to control inflation. Then, the economic growth after the crisis
caused an important increment in the natural gas demand. At the same time, losing the
price signal discouraged investments in exploration. Hence, demand exerted pressure on
a weak supply on the Argentinean side. In 2004 the Argentinean government decided to
suspend exports of excess supply of natural gas in order to keep the internal demand
satisfied (SECRETARIA DE ENERGIA, 2004). The gas crisis started, peaking in 2007
when Argentinean exports became null.
The main input in thermal power generation in Chile until 2003 was the
Argentinean natural gas. Combined cycles power plants had become the main
technology for the expansion of the power sector, reaching 36% of total electricity
produced in 2003. At that point, the gas crisis had an important effect on Chile.
MUÑOZ et al. (2004) calculated how large the costs for Chile due to the Crisis were.
According to MUÑOZ et al. (2004), the cost of moving from natural gas to coal for the
main power system in Chile was USD 353 million per year. Such cost rose final power
prices in 13%.
After 2004, the reduction in Argentine exports was compensated by more coal-
based generation, but by 2006 this was not enough. Diesel and fuel oil had to be used to
cover the difference, resulting in a relevant increase in spot prices, at the beginning, and
later on contract prices.
In front of this situation, the Chilean government decided to proceed in two
venues. First, a new legislation was enacted to incentivize investment on new power
plants to replace combined cycles by replacing the regulated tariff system for contract
prices with an auction on long-term contracts. The second measure was to establish
LNG terminals.
During these milestones, Bolivia played a key role as a gas supplier to Argentina
and also to Brazil. The role of Bolivia as agas-exporter country dates back to the 1970s
with the 441-km YABOG pipeline construction to supply gas to Argentina. The long-
term contract with Argentina to supply gas through YABOG expired in 1992, but it was
extended to 1999 so Argentina would support the incomes of Bolivia until the start of
29
the exports to Brazil. Over the life of this period, Bolivia exported almost 50 bcm of
natural gas to Argentina, worth about US$4.3 billion (in nominal dollars) for an average
price of $2.21/mmbtu (VARGAS SALGUEIRO, 1996).
In a search for an outlet of Bolivian gas resources, in 1996 YPFB signed a
definitive contract for the sale and purchase of natural gas and a contract for pre-
payment with Petrobras. This gas supply agreement included ramping volumes and
take-or-pay commitments. To transport the gas, the construction of the 3150-km
GASBOL pipeline began in 1997 and was completed in 1999. The market guaranteed
by Petrobras, encouraged massive investments in exploration and pipeline construction
in Bolivia, which in turn produced certified (proven and probable) natural gas reserves
that increased from a level around 170 bcm before 1997 to 1 481 bcm by 2001
(MARES, 2006). Imports from Bolivia represented 26% of natural gas supply for Brazil
in 2000, just one year after GASBOL operations started in 2005 this number increased
to 50%. By that time, in 2004, when the Argentina-Chile gas crisis started, rather than
raise domestic prices, the Argentine government turned to Bolivian exports to relieve
Argentine shortages in 2004 (MARES and MARTIN, 2012).
In 2006, when Evo Morales’ administration began, the government nationalized
all petroleum operations, proclaiming the ownership of the oil and gas industry and
forcing operators into joint venture contracts with Yacimientos Petrolíferos Fiscales
Bolivianos (YPFB) (CHAVEZ-RODRIGUEZ et al., 2016a). In that year, the Bolivian
government, in a renegotiation contract strategy with Petrobras, threatened to cut off the
gas supplies to Brazil, exposing the weakness of the Brazilian natural gas supply chain
(MARTINS et al., 2016).
Also in 2006, a contract on interruptible basis between YPFB and ENARSA, the
gas national company of Argentina, with duration of 21 years was subscribed. It began
with an initial volume of 7.7 MM3/day with possible increases of up to 27.7 Mm3/d
over 10 years. Moreover, in 2012, YPFB and ENARSA signed a new interruptible
contract with maximum volumes of 2.7 Mm3/d.
However, during the fall of 2007 Bolivia, having failed to attract the necessary
investment to produce all the natural gas it contracted for the export and domestic
30
markets, reduced its supply to Argentina by over 50% in order to fulfil its contracts with
Brazil, which was not only a bigger market but also threatened to invoke penalty clauses
for failure to deliver (the Argentine government said it would not penalise Bolivia)
(MARES and MARTIN, 2012).
In the last years, exports from Bolivia to Argentina and Brazil were regular.
Bolivia successfully increased its production. However, in 2016, with the risk that
YPFB would not be able to deliver the volumes committed by contract with ENARSA,
Argentina had to set up an agreement to import natural gas from Chile, using the
overcapacity of this last one.
The milestone, projects and cutting-off threats explained in this section regarding
the natural gas exchanges in the Southern Cone depicts the relevance of gas flows by
pipeline for these countries. In this sense, considering the development of
unconventional gas in Argentina and gas associated in Brazil but also with an increasing
gas demand, will regional natural gas trade in the Southern Cone increase in the next
years?.
2.3.4 The LNG expansion in the Southern Cone
Natural gas demand in Argentina, and in Chile to a lesser degree, is strongly
affected by variations in temperature during the year (the coldest months in the
Southern Hemisphere are from May to September). Moreover, as shown inFigure 2.3,
during the scarcity years of natural gas in Chile, the buildings sector was prioritized in
detriment of power generation. In Argentina, to balance the natural gas supply-demand,
thermal power is rationed during the winter prioritizing the provision of natural gas for
the buildings sector (RODRIGUES, G. M. S.; OLIVEIRA, 2015).
31
Figure 2.3. Historical natural gas consumption in Chile and Argentina in
Buildings (residential, commercial and public sector) and power generation sectors
between 2006 and 2011. Source: CNE (2016c) and ENARGAS (2015)
As explained in the previous section, due to the scarcity of natural gas, LNG
regasification units were installed in Argentina and Chile. Actually, in 2008,in Bahia
Blanca-Argentina, the Floating LNG Regasification Unit (FRSU) started operation with
a nominal capacity of 10 Mm3/d (RODRÍGUEZ, 2011) loaded with around 36,000
tonnes from Trinidad and Tobago. Interesting enough, as it was the first project of this
nature in the region many security and environmental concerns were raised by that time
(LA NACION, 2008; NUESTROMAR, 2008). This was the second operation ship-to-
ship performed in the world after the one in Teeside, United Kingdom (SCHNEIDER,
2008). An illustrative picture of this operation is shown in Figure 2.4. In those years, the
LNG regasification project in Argentina, promoted by YPF and executed by ENARSA,
was justified by ensuring supply to face hypothetical disruptions of Bolivian natural gas
imports(DICCO, 2011). The regasification capacity of this project was expanded to 17
Mm3/d (YPF, 2015). A second FRSU plant in Argentina, Puerto Escobar, started
operation in 2011, with a regasification capacity of 17 Mm3/d (RODRIGUEZ, 2011).
32
Figure 2.4. A ship-to-ship transfer of LNG in Bahia Blanca Gas Port.
Source: EXCELERATE ENERGY (2016)
In Chile, the first onshore terminal in the central region of the country (Quinteros)
was carried on by ENAP, the public oil company, BG Group, Enagás, Endesa Chile and
Metrogas. It started operation in October of 2009 with a nominal capacity of 10 Mm3/d.
In 2015 is was expanded by 5 Mm3/d (GNL QUINTERO, 2015). To supply power
plants for mining activities in the North of Chile, a second terminal was installed in
Mejillones by CODELCO, the public copper company, and GDF SUEZ Energy Chile in
2010, with regasification capacity of 5.5 Mm3/d (CABANES, 2015). This terminal has
an onshore regasification train and an LNG floating storage unit
(FSU)(HYDROCARBONS-TECHNOLOGY, 2008). In addition, the current
government has defined an Energy Agenda for 2014-2018, where one of the goals is to
promote the use of LNG to reduce diesel use and, consequently, spot prices in the power
market.
In the case of Brazil, the risks of disruptions of Bolivian gas imports of natural gas
also triggered the development of LNG as an alternative source to increase gas supplies
and to ensure the continued provision of this energy source (MARTINS et al., 2016).
33
There are currently three LNG regasification offshore terminals in Brazil. One,
known as Pecém, is located on the northeast coast in the state of Ceará, with a
regasification capacity of 7Mm3/d. Another terminal is located on the Guanabara Bay,
on the coast of Rio de Janeiro, with a capacity of 20.7Mm3/d. In January 2014, a third
terminal began to operate in Baía de Todos os Santos, Bahia, with a capacity of
17Mm3/d. LNG imports in Brazil played a key role in the recent drought years.
LNG is increasing its share in the natural gas supply in the Southern Cone and
prices have been declining during the last years, but, will LNG imports continue to
increase?, Can the Southern Cone become an LNG exporting region?. In a
comprehensive way: “What role LNG will play in the natural gas market of the
Southern Cone?”.
.
2.3.5 The Hydropower in Brazil and Wind and Solar potential in the Southern Cone
In 2015, electricity produced by hydropower plants accounted for 64% of the total
electricity produced in Brazil(EPE, 2016b). Up to 2024 more 27.4 GW of hydro power
plants are planned to be added to Brazil’s power generation system(EPE, 2015b). This
is part of the Brazilian government actions, along with private investors, in expanding
hydropower generation in Amazon, where about 30 large dams are planned to be
constructed in the next 30 years (Figure 2.5).
34
Figure 2.5. Map of Brazilian Amazon showing existing and planned hydroelectric
power plants. Source: ALMEIDA PRADO JR. et al. (2016)
For these new hydropower plants, there is a strong trend towards building run-of-
the-river plants, using ever-smaller reservoirs with a good ratio between installed
capacity and the area to be flooded (ANDRADE and DOS SANTOS, 2015). As the
Amazon is an ecologically sensitive area, the run-of-the-river configuration intends to
lower environmental impacts, however, it leads to a lower energy storage capacity,
which is necessary for critical drought periods, and to a lower floodwater storage
capacity in flooding period.
In 2013, Brazilian reservoirs reached the lowest levels since 2001 (ONS, 2016).
Despite the internal transfer of energy between regions, it was necessary to turn on
thermal plants to compensate for the lower generation from hydroelectric plants
(ALMEIDA et al., 2016). The capacity factors of hydroelectric plants have been
decreasing in Brazil in the last years as shown in Figure 2.6. Differently, from other
regions in the country, the dry months in the North Region are June to December).
Therefore, the expansion of hydropower in the Amazon is likely to increase the
35
Brazilian energy system linkage to fluctuations in rainfall and natural gas-fired power
plants can be a key source of reliability in electricity supply.
Figure 2.6. Historical capacity factor between 2012 and 2015 for hydropower
plants in Brazil. Source: ONS (2016)
The Southern Cone has seen significant investment in non-hydro renewable
energy as well. In recent years, Brazil had accumulated investments of 75.3 billion US$
from 2009 to 2014, followed by Chile and Argentina with 8.5 and 1.8 billion US$
respectively for the same period (BNEF, 2016). For both Argentina and Brazil, wind
power is leading the major investments whereas in Chile, solar power is the main driver.
This obeys to the resources and investments suitability, as can be observed in Figure
2.7.
36
Figure 2.7. Investment suitability for (a) Wind power and (b) Solar power in
South America. Source: IRENA (2016)
It is projected in different expansion plans that wind and solar technologies are
going to increase their share dramatically in the following years in the Southern Cone
(AGEERA, 2012; CNE, 2015; EPE, 2015b; MHE, 2012). However, this increase of
stochastic and variable renewable energy will bring challenges to the design and
operation of the energy systems. As suggested by SCHMIDT et al. (2016), the
intermittence of these technologies causes problems in transmission grid and increases
the need for quickly ramping backup capacities, such as flexible natural gas power
plants; based on this natural gas power capacity will be required to deal with the
expansion of wind and solar power in the Southern Cone.
As observed in Figure 1.2, the power generation is the sector with most
consumption of natural gas in all countries of the Southern Cone. Renewables are
increasing its share at a fast pace in this region. In this sense, “will renewable energy
jeopardize the natural gas consumption in power generation?”.
These issues will be addressed in this thesis in Chapter 6 in a general scope, and
in Chapter 7 each of these topics is discussed analysing its role in the natural gas
markets of the Southern Cone.
37
3 Methodological Procedure
This chapter explains the methodological procedure adopted to attain the
objectives proposed in this thesis. The rationale behind the research will be described
while focusing on the modelling steps that were adopted and, in some cases, developed
in order to build the demand and supply of natural gas and power generation. Finally,
data sources and information gaps will be detailed.
3.1 Research Design
3.1.1 Concept
The approach adopted to explain the research questions and attain the objectives
of this thesis is based on scientific modelling, defined as the symbolic representation of
the phenomena from which one may compute non-trivial inferences (SMITH, T. R. et
al., 1995). In scientific modelling, an open-ended set of concepts that characterizes
significant aspects of the phenomenon is represented, which is, in our case, the supply
and demand in natural gas markets.
Since the real dynamics of natural gas markets is based on a wide range of
concepts, including human behaviour, the organization of the market and politics, which
will not be addressed by this thesis, it is important to define the set of concepts where
our symbolic representation of the phenomena relies on. The concepts chosen for the
scientific modelling to represent the natural gas markets adopted in this study are the
following:
- Conservation of mass and energy.
- Perfect competition in the market, which include symmetric information
between the agents and lack of market and government failure.
- Perfect foresight.
38
The objective function adopted in the model is the minimization of the total
system costs, which is fully equivalent to the maximization of total surplus or profits
(LOLOU et al., 2004). In a perfect competition market, the maximization of profits of
each economic agent will result in a total maximization surplus of the society
(SAMUELSON, 1952).
The equilibrium of the market is reached when the total surplus (or minimized the
total cost) of the system is maximized as shown in Figure 3.1.
Figure 3.1. Equilibrium of the supply and an elastic demand of a specific commodity.
Source: LOLOU et al. (2004)
Since the natural gas markets are addressed in this study under a microeconomic
perspective, the model used computes a partial equilibrium on the natural gas and power
sector. The flows of energy and materials, as well as their prices, are computed by the
model in a way that the suppliers of energy produce exactly the amounts that the
39
consumers are willing to buy only in these markets (LOLOU et al., 2004), without
taking into account the general equilibrium of the whole economy.
As the key factors discussed in Chapter 2 are on the supply side, and due to data
limitation, for the countries studied, to construct a bottom-up demand based on energy
services, the natural gas demand for final end-uses and electricity demand were
modelled as fixed values, e.g. as perfectly inelastic (Figure 3.2).
Figure 3.2. Equilibrium of the supply and a fixed demand of a specific commodity.
Source: LOLOU et al. (2004)
Despite the fact that the electricity demand modelling is fixed, the use of natural
gas for power generation has an elastic behaviour as it competes with other fuels and
uses in different technologies; hence, when natural gas has a high cost the natural gas
demand for power generation is lower and vice versa. Consequently, the composite
curve of the natural gas demand results from the sum of a fixed (or perfectly inelastic)
demand for end-uses and an elastic demand for power generation.
On the supply side, the model should reflect the variety of technological options
that investors have for production, processing and transportation of natural gas and the
technologies that consume natural gas in the power generation sector. Costs used to
40
model the different technologies in this study represent estimates at the concept level of
the project based on the literature. These costs might significantly differ from the costs
on the feasibility or engineering level. In addition, there is evidence that the CAPEX for
oil and gas technologies fluctuate according to the oil price (ALOSHBAN and
SANDREA, 2006; VIEGAS, 2013), and also costs learning curves alter the feasibility
of some technologies, like what happened for the unconventional gas technologies in
the last decade (FREDD et al., 2015). For the sake of simplicity, the CAPEX applied for
the different technologies will remain constant over the assessed horizon. Finally, a real
discount rate of 10% per year is incorporated to assess the investments as this is a
standard value in the oil and gas industry (HARDEN, 2014; SMITH, J. L., 2015; SRR,
2015).
The time horizon for the model is chosen based on the logic of long-term strategic
planning of the major energy companies and governments in South America, which is
between 2025 and 2030 (AGEERA, 2012; CNDC, 2014; ECOPETROL, 2016; ENAP,
2014; EPE, 2015b; PDVSA, 2015; PETROBRAS, 2014a; UPME, 2012). Many energy
companies have investments and business plans for a 5-year period (ECOPETROL,
2015; PETROBRAS, 2014a; YPF, 2013; YPFB, 2015a). However, the key factors that
will be assessed, such as renewable power generation plants, the development of
unconventional gas in Argentina and the development of associated gas of the Brazilian
Pre-Salt fields, require more time to reach full implementation. Consequently, applying
a time horizon until 2030 is appropriate for the objective of this thesis.
The purpose of this scientific modelling approach is not to predict but to explore
the role of the key factors discussed in Section 7 to shape the future of natural gas
markets under different scenarios. The quantitative results obtained from the models
will be further discussed taking into account the assumptions and limitations of this
modelling exercise. Moreover, the results of this scientific modelling can be used to
identify policy opportunities for governments to drive their own energy markets as close
as possible to a maximization of total surplus, under the constraints of their own market
structure and national goals.
41
3.1.2 Modelling Approach
A hybrid approach was developed to project the natural gas supply and
consumption, by combining a simulation model (LEAP - Long range Energy
Alternatives Planning System) on the demand side, with a technological rich energy
system optimization model (TIMES) on the supply side (Figure 3.3). The TIMES model
was the centerpiece of the modelling approach, being fed with inputs from: i) LEAP
regarding the end-use demand of natural gas per sector; ii) a Multi-Hubbert model to
project oil production and estimate associated natural gas production, given the oil-to-
gas ratios; iii) an excel-based model for the projection of non-associated natural gas
production based on the resources (proven, probable and possible for both conventional
and unconventional types), estimating the production costs based on investments; iv)
assumptions from the national electricity generation expansion plans regarding the
evolution of the annual electricity demand and the annually installed capacity, and also
the historical capacity factor monthly-curves for different technologies by regions; and
v) the exogenous estimation of installed capacities and costs of natural gas technologies.
Figure 3.3 illustrates the interactions between these components.
42
Figure 3.3. Natural Gas & Power modelling approach developed for this study
As depicted in Figure 3.3, LEAP provides the natural gas demand for end-uses
(residential, commercial and public, transport and industrial) as fixed values to TIMES.
Natural gas for power generation is modelled based on the power expansion
plans(capacities by technologies) and electricity demand projections adopted for each
country. The natural gas supply chain was modelled in TIMES. In the case of upstream,
to model non-associated gas production a supporting algorithm was developed based on
reserves and resources classification (CHAVEZ-RODRIGUEZ et al., 2016a) and then
inputted into the TIMES model. Associated gas was modelled externally using a Multi-
Hubbert approach and inputted as fixed values according to the estimated ultimate
recovery (EUR) used. These methods are described in further detail in the following
sections.
43
Provided these inputs, the constructed architecture in the TIMES model (here
thereafter referred to as TIMES-ConoSur) performs:
- The least-costs operation of power generation, estimating the natural gas
demand for this sector.
- Electricity trade between regions via international transmission lines.
- Natural gas trade between regions via international gas pipelines
- The interplay between the consumption of natural gas for electricity
generation and of that for end-use consumption (i.e. residential, commercial,
transport and industry).
- The interplay between the domestic production of natural gas and the imports
of this commodity.
- The projection of new capacities and the production of non-associated
natural gas, incorporating the shadow price effects in the objective function
of condensates and NGL production.
- The LNG trade using regasification plants or liquefaction plants for imports
and exports, respectively.
- The expansion and operation of existing and new Natural Gas Processing
Plants (NGPP).
- The least cost for the system regarding the associated gas production
monetization options.
It is relevant to highlight that the capacity expansion of power plants by
technologies is constrained by the plans exogenously inputted in TIMES-ConoSur.
Consequently, the model does not expand capacities for power generation and it is
limited to the least-cost operation in the electricity sector. Therefore, TIMES-ConoSur
is basically a natural gas model for the Southern Cone, able to provide the least cost
solution for expanding and operating the gas industry in Southern Cone countries, given
a previously defined electric power capacity.
Finally, TIMES-ConoSur provides the natural gas demand by end-uses and
power generation, the domestic natural gas production curves and costs, the trade of
natural gas by either LNG or international pipelines, the capacity expansion of
technologies, CAPEX and OPEX costs, and total energy system costs.
44
3.1.3 Geographical Resolution
The geographical scope of the model covers four countries: Argentina, Bolivia,
Brazil, and Chile. The selection of these countries is based on their major role either as
suppliers or consumers in the Southern Cone, but even more important, it is necessary
to understand the natural gas dynamics in this region which is the most integrated
natural gas market in the continent due to existing international pipeline system. As
explained in Chapter 2, there is a strong dependence on Bolivian natural gas in
Argentina and Chile, but also, provided Argentina can raise their domestic natural gas
production, this country can re-take its exports to Chile and Brazil. Alternatively, as
happens nowadays, Chile is exporting natural gas to Argentina during its winter peak
consumption using LNG regasification and bi-directional pipelines. This first
geographical approach intends to address these issues based on the existent
infrastructure and the capacity for natural gas trading within the Southern Cone. Further
studies can be done using this methodology for the whole South America.
In this study, Chile and Brazil were split both into two regions: “North Chile” and
“Central-South Chile”; “Integrated Brazil” and “North Brazil”. This division was based
on the independence of gas pipeline networks, the different sources of natural gas
supply and the different structure of the natural gas demand.
Additionally, in the case of Chile, the power grids of “North Region” and
“Central-South Chile” are not yet interconnected. Moreover, the natural gas
consumption in North Chile is mainly for power generation to supply electricity for the
mining sector, in contrast with Central-South Chile, where natural gas is used mostly for
power generation but is also relevant for the industrial and buildings sectors.
In the case of Brazil, EPE (“Empresa de Pesquisa Energética”) make the same
disaggregation for natural gas markets in their Ten-year Energy Expansion Plan (EPE,
2015b): “Malha Integrada” (Integrated Network) and “Região Norte” (Northern
Region). There are differences in the structure of the demand between these two regions
as well, but the options for natural gas supply sources are even more relevant. While in
“Integrated Brazil” the market of natural gas has access to offshore/onshore fields
45
associated and non-associated gas, Bolivian imports, and LNG, in the “Northern Brazil”
the market has only access to onshore associated and non-associated gas production in
the Amazon State and a limited supply of Bolivian gas for a thermal power plant in
Cuiaba, Mato Grosso.
The natural gas trade through international gas pipelines and electricity through
transmission lines using this geographical resolution is also modelled in TIMES-
ConoSur, as shown in Figure 3.4.
Figure 3.4. Geographical Resolution of the TIMES-ConoSur model
3.1.4 TemporalResolution
It is a usual practice to model the temporal resolution in seasons and three time
periods by day: Day, Night and Peak (SIMOES et al., 2013). However, the
characteristics of natural gas markets at the different stages of the supply chain require a
particular approach.
46
On an intra-annual level, different resolutions have been experimented with the
aim of finding a good trade-off between detailed temporal representation and model
manageability.
Detailed intra-annual representation is relevant in particular for the case of natural
gas for several reasons:
i) Flow commitments in contracts are often settled on a daily basis
(SUTHERLAND, 1993),
ii) Gas power plants also play an important complementary role with
intermittent unpredictable renewable power plants as wind and PV
(KANNAN, 2011).
iii) There is seasonality of natural gas demand for final sectors.
A first hourly-resolution approach was tested in (CHAVEZ-RODRIGUEZ et al.,
2016b). The outcomes using this time resolution are interesting on the demand side
where it is possible to observe the behavior of natural gas consumption for power
generation throughout the day and a significant increase in the peak hours. However,
without the condition to model the storage capacity of line-packs in pipelines, an hourly
time resolution fails to represent a real-world behaviour to balance natural gas demand
consumed in the markets and the production injected in the pipeline. Therefore it was
necessary to use a more aggregated temporal resolution.
Based on these considerations, a monthly resolution for modelling natural gas for
both midstream technologies and upstream technologies was adopted (12 months per
year). This temporal resolution was used also to input the exogenous end-uses demand.
However, for power generation was used a temporal resolution of 24 hours per typical
day of the month: 24 x 12 = 288 time-slices per year. This temporal resolution for
power generation matters for the consumption of natural gas since natural gas-fired
power plants operate during the peak hours of the day as shown further on.
47
3.2 Tools description
As mentioned before, the main software tools used for this study were LEAP, to
elaborate the natural gas demand, and TIMES as the integrator and optimization energy
tool. Their features are described in the following sections.
3.2.1 LEAP
LEAP is a scenario-based energy-environment modelling tool for the analysis of
energy policy. LEAP can be used to track energy consumption, production and resource
extraction in all sectors of an economy(KEMAUSUOR; NYGAARD; MACKENZIE,
2015). LEAP uses a bottom-up technological and sectorial approach on the level of
energy services or the level of final demand, according to the availability of data.
However, top-down approaches can be inputted in LEAP exogenously.
LEAP is designed around the concept of long-range scenario analysis. Scenarios
are self-consistent storylines of how an energy system might evolve over time. These
scenarios can be used to describe individual policy measures, which can then be
combined in different options into alternative integrated scenarios, for instance, a
scenario of high-penetration of natural gas. This approach allows policymakers to assess
the marginal impact of an individual policy as well as the interactions that occur when
multiple policies and measures are combined (COMMEND, 2016).
The model was developed by the SEI-US (Stockholm Environment Institute),
based in Boston, Massachusetts (HEAPS, 2012). LEAP can be applied at different
scales ranging fom cities and states to national, regional and global applications
(SUGANTHI and SAMUEL, 2012).
LEAP has been widely used in Latin America. There are several reasons for its
popularity in the region, and also for why it was adopted to project the demand of
natural gas in this thesis:
- The demand is based on the logic of scenarios, which are self-consistent
storylines of how the energy system might evolve over time.
48
- Low initial data requirements are a key feature for Latin America, where
primary information to construct energy models can be scarce for some
countries.
- Simple and transparent accounting (similar to an excel sheet) – see Figure
3.5.
- Low learning barriers
- Friendly interface
- It is freely available for a relevant period of time for scientific research
and government agencies in developing countries.
Figure 3.5. Example of LEAP’s interface. Source: HEAPS (2012)
3.2.2 TIMES
The MARKET ALocation (MARKAL) model is a widely applied bottom-up and
dynamic linear programming model developed by the ETSAP (Energy Technology
Systems Analysis Programme) of the IEA (International Energy Agency). MARKAL is
49
a generic model tailored by the input data to represent the evolution over long-term
periods, usually 40–50 years, of a specific energy system on a national, regional, state
or city level (AMORIM et al., 2014).
The Integrated MARKAL-EFOM system (TIMES) is an evolved version of
MARKAL and of the Energy Flow Optimisation Model (EFOM) with new functions
and flexibilities, also developed within the ETSAP (LOLOU et al., 2005). In both
MARKAL and TIMES the objective function is the minimization of the net present
value of the total system costs over all regions and periods, including the “costs” of lost
demand (welfare loss), as shown in Eq. 3.1:
(∑ ∑ ( )
) Eq. 3.1
where:
NPV is the Net Present Value of the total costs for all regions over the
whole time horizon;
is the general discount rate (10%);
REFYR is the reference year for discounting (2012);
YEARS is the set of years in which costs are incurred (2012…2030);
R is the set of regions in the scope of the study (in this study: 6
regions);
ANNCOSTS (r,y) is the total annual cost in region r and year y;
This cost minimization objective equals the total surplus maximization, as discussed
in Section 3.1.1. The following costs categories were taken into account for the
minimization of system costs in this study:
- Capital costs, incurred for investing (in our modelling we did not
consider decommissioning capital costs).
50
- Fixed and variable annual operation and maintenance costs.
- Costs incurred for exogenous imports (out of the Southern Cone) and for
the production of domestic resources.
- Revenues from exogenous exports (exporting to countries out of the
Southern Cone).
- Delivery costs for required commodities consumed by the process.
- Salvage value of processes and embedded commodities at the end of the
planning.
There are other category costs that were not included in the objective function of
this study:
- Taxes and subsidies associated with commodity flows and process
activities or investments (that for natural gas would be for instance the
government-take in the upstream).
- Welfare loss resulting from reduced end-use demands (as the demand
modelled in this study is perfectly inelastic, the welfare loss was
neglected).
The main advantage of TIMES in comparison to its predecessors is its flexibility
since it is possible to sub-divide the year into several time periods with different user-
defined lengths. Among other features, this enabled running a first hourly model of
natural gas (CHAVEZ-RODRIGUEZ et al., 2016b). TIMES-based models are built on
linear programming (LP), but can also include mixed integer programming (MIP) or
multi-stage stochastic programming (LOLOU and LEHTILA, 2012). MIP is used for
instance to expand a technology with typical sizes (for instance, a nuclear plant of 1
GW). Also, it is possible to have different levels of disaggregation for different sectors
and the option of making an investment in blocks. The main output of TIMES are
energy system configurations, which meet the end-use energy service demands (in our
case final energy demand) at least cost while also adhering to the various constraints
(e.g. a restriction of CAPEX investments in gas upstream). The model outputs are
energy flows, energy commodity prices, GHG emissions, capacities of technologies,
energy costs and the abatement costs of the marginal emissions (LOLOU et al., 2005).
In Figure 3.6, a schematic of the TIMES model is shown, along with outgoing white
block arrows that show the model outputs.
51
Figure 3.6. Schematic of TIMES inputs and outputs. Source: REMME et al.
(2001)
The TIMES energy economy consists of three types of entities (LOLOU et al.,
2005):
- Processes are representations of physical devices that transform
commodities into other commodities. Processes may be primary sources
of commodities (e.g. gas well, LNG imports), or transformation activities
such as conversion plants that produce electricity, energy-processing
plants such as natural gas processing plants, end-use demand devices
such as cars and heating systems, etc.
- Commodities consist of energy carriers, energy services, materials,
monetary flows, and emissions. A commodity is generally produced by
some process(es) and/or consumed by other process(es).
- Commodity flows are the links between processes and commodities. A
flow has the same nature as a commodity but is attached to a particular
process, and represents an input or an output of that process.
52
TIMES is a model generator that, based on the input information provided by the
modeller, generates an instance of a model. In this study, TIMES worked as the model
generator of the natural gas model & power, here named as TIMES-ConoSur.
Despite its non-intuitive interface and significant learning efforts, TIMES is widely
used across the world: in nearly 70 countries (IEA-ETSAP, 2016). However, in Latin
America it is rarely used. This thesis and the work of POSTIC (2015) may be the first
regional studies using TIMES for South America on a significant geographical
disaggregation level.
53
4 Natural Gas Demand for End-uses Modelling
As explained in Section 3.1.1, due to data limitation for building a bottom-up
demand at the energy services level or with elasticity coefficients, the natural gas
demand for end-uses was elaborated as a perfectly inelastic demand at the final energy
level. This means that any change in the costs of natural gas does not affect the total
final demand of natural gas in end-uses.
Forecasting techniques for the natural gas demand are predominantly top-down
approaches (SOLDO, 2012). However, bottom-up approaches provide more details
about the structure of the demand, allowing to background the projections with
storylines and to test energy policies (e.g. energy efficiency measures, fuel substitution
programs, etc.) (CHAVEZ-RODRÍGUEZ et al., 2014).
Due to the heterogeneity of the final energy use of natural gas in Southern-Cone
countries and the lack of data for a detailed standard bottom-up approach, two basic
methodologies were applied: one based on a parametric simulation using the software
Long Range Energy Alternatives Planning System (LEAP); the other based on an
econometric fit, aiming at deriving natural gas demand functions. LEAP was used to
build natural gas demand projections in residential (household) and transportation
sectors through2030. Industrial and commercial/public sectors were initially modelled
through an econometric approach and then the results were inserted into LEAP for
accounting. Figure 4.1 shows the hybrid approach adopted to project the natural gas
demand for end-uses.
54
Figure 4.1. Methodological procedure for the elaboration of natural gas demands
for end-uses.
In the case of Brazil, the natural gas demand for end-uses was not projected but
instead the natural gas projections of EPE (2014, 2015b) were adopted. In the following
sections, the modelling methodology applied to project the natural gas demand by
sectors in Argentina, Bolivia and Chile are shown.
In the case of Chile, as stated in Section 3.1.3, the country was divided into two
regions: North Chile and Central-South Chile. Table 10.1 in Annex A shows the base-
year calibration to split the natural gas demand for end-uses into these two regions.
As the natural gas demand projections for end-uses in Chile was estimated in
LEAP at the national level, potential market drivers such as a number of households,
vehicles, etc. were used to split the increments of demand between North Chile and
Central-South Chile.
55
4.1.1 Residential sector
This study adopted a bottom-up approach to model the natural gas demand in the
residential sector. A generic equation of this procedure is shown in Eq. 4.1.
serviceenergyj
jijijii ISAE ,,,
Eq. 4.1
where i is the fuel projected, j is energy service (cooking, heating, etc.), Ei is the total
fuel i consumed in the residential sector, A is the activity (number of households), S is
the share of the activity using the fuel i to attend the energy service analysed, and I is
the specific energy consumption of fuel i to attend the energy service j (i.e. m3 of
natural gas per household per year for cooking).
Household projections based on average occupancies per household and
population projection from UN-DESA (2012) for Bolivia and Chile, and the INDEC
(2015) projections for Argentina, were used as a driver of the energy consumption in the
residential sector. In the case of Argentina, Bolivia and Chile, 3.6, 3.3, and 3.5 persons
per household ratios were used respectively, based on their national households’
surveys. Although these ratios may change in the time horizon adopted in this study,
performing a detailed demographic forecast is outside the scope of this work.
The consumption of natural gas in the residential sector was calculated based on
three main energy services: cooking, water heating and space heating. For each of these
uses, the market share of natural gas was estimated for the base year. The most reliable
information to estimate this market share was the national household surveys, which
showed shares of fuels used for cooking purposes. It is assumed that there is a
proportional relation between shares by fuels for cooking and other end-use. Then,
further projections were made on the growth of the natural gas market share over the
56
time horizon of this study. The specific consumption of natural gas per household was
assumed to remain unaltered for all energy services.
For Argentina, the end-use energy consumption by fuel type in households at the
country-level was adjusted using data from National Population, Households, and
Dwelling Census of 2010 (INDEC, 2012) and the National Energy Balance of 2012
(SECRETARIA DE ENERGIA, 2015a). Data for monthly natural gas consumption and
the number of customers in the residential sector from ENARGAS (ENARGAS, 2015)
were analysed in order to estimate specific consumption for end-uses in households. As
depicted by many authors (FILIPPÍN et al., 2011, 2012; GIL, 2007; GONZÁLEZ et al.,
2007; VAGGE and CZAJKOWSKI; CELINA, 2008), natural gas consumption in the
residential sector in Argentina is highly explained by the consumption for space heating,
which has a seasonal behaviour and depends on external temperatures. For modelling
purposes, we estimated an average monthly curve of natural gas consumption based on
the historical consumption from 1993 to 2014 (ENARGAS, 2015), and by the criteria
developed by GIL (2007) we separated the natural gas consumption for heating from
other uses. Not only for Argentina, but also for Bolivia and Chile, the specific
consumptions for cooking and water heating used by CHAVEZ-RODRÍGUEZ,
(2014)were taken as reference.
In the case of Bolivia, the national household survey (INE, 2013a), data from
(MHE, 2013) and proxys for energy consumption for energy services based on
CHAVEZ-RODRÍGUEZ (2014) were used to estimate the specific energy consumption
for energy services by source. In this estimation, natural gas consumption in other uses
– such as space and water heating – was estimated to reach the same level reported in
the National Energy Balance for the residential sector – 7,090 kboe (MHE, 2013).
As observed in the monthly consumption of natural gas in the Bolivian residential
sector (ANH, 2015), there is not a seasonal pattern that could be attributed to heating
use. Based on this, the natural gas consumption in Bolivia for end-uses was maintained
uniform throughout the year.
In the case of Chile, the energy consumption in the residential sector was
modelled based on data from the National Population and Households Census of 2002
57
(INE, 2003), Results of Census 2012 (INE, 2013b) and the National Energy Balance
(MINENERGIA, 2013). In Chile, the same pattern of seasonal consumption of space
heating is shown in the monthly consumption statistics of LPG and natural gas (CNE,
2016c). We applied the same methodology used for Argentina in the Chilean residential
sector based on historical data of CNE (2016c). Figure 4.2 presents the yearly curve of
natural gas consumption in the modelled households. The higher consumption in
Argentina may be explained by the overspending of natural gas caused by the subsidies
as evidenced by (GIL, 2013).
Figure 4.2. Natural gas consumption curve estimated for households by end use
for Chile and Argentina. Source: Own elaboration based on ENARGAS (2015) and
CNE (2016c).
Table 4.1 shows the increase in the natural gas share for cooking purposes in
households considered in the projection. The expansion of natural gas use in
Argentinian households represents efforts to continue expanding connections at
historical rates (1.2% per year). Reaching 80% of natural gas penetration in households
would be possible with the construction of the Argentina Northeast Gas Pipeline
(GNEA), a 4131 km gas pipeline network with a capacity of 11.2 Mm3/d to supply
natural gas to Salta, Formosa, Chaco, Corrientes, Misiones and Santa Fe provinces
(ENARSA, 2015).
58
In the case of Bolivia, the assumed growth of natural gas was qualitatively
supported by the efforts of the government to expand the distribution network and
household connections. In 2014, the YPFB investments in the primary and secondary
gas network infrastructure reached 158 million US dollars with an accumulated amount
of 566 million US dollars since 2006 (YPFB, 2015d). The number of domestic
connections rose almost nine-fold – from 13,800 in 2007 up to 122,000 at the end of
2014 (YPFB, 2015d).
In Chile, since the gas transport and the distribution systems are atan early stage
and have a higher potential to grow, we have considered a rate of growth for
connections of 5.6% per year (which is conservative compared to the Bolivian historical
growth rates). This assumption is aligned with the Strategic Plan for 2014-2015 of the
State Oil Company, ENAP (ENAP, 2014), which considers boosting natural gas for
power and residential consumption as an axis of its plan.
Table 4.1.Share of households for the cooking energy service.
Argentina
Fuel 2001a 2010
a 2030
b
Firewood and other biomass 5% 3% 1%
Natural gas 57% 63% 80%
LPG 38% 34% 19%
Others 0% 0% 0%
Bolivia
Fuel 2001a 2012
a 2030
b
Firewood and other biomass 38% 26% 10%
Natural gas 0% 10% 35%
LPG 58% 61% 53%
Electricity 1% 1% 1%
Others 1% 0% 0%
No cooking 2% 2% 2%
Chile
Fuel 2002a 2012
a 2030
b
Firewood and other Biomass 12% 10% 5%
Natural gas 9% 11% 30%
LPG 78% 79% 64%
Others 1% 1% 1%
Source: a. Census data available (INDEC, 2003, 2012) for Argentina, (INE, 2013a) for Bolivia and
(INE, 2003, 2013b) for Chile
b. Author’s projections.
59
4.1.2 Commercial and Public Sectors
Since publically available and reliable data for the commercial/public sectors in
the assessed countries is too limited to perform a bottom-up approach, this study
adopted an econometric approach. As the natural gas distribution grid serves both
households and commercial buildings, this study considers that the number of
commercial users connected to the distribution pipelines follows the same geographical
expansion pattern of the residential sector, as indicated in Eq. 4.2:
(4.2)
Where and are the number of users connected to distribution pipelines
in the commercial/public and residential sectors, respectively and and are
coefficients. Then, the average consumption of natural gas in the commercial/public
( ) and residential sectors ( ) (in kboe) were calculated as following:
(4.3)
(4.4)
The variables and stand for the specific consumption of natural gas per
user in the commercial/public and residential sectors3, respectively (in kboe/user).
Finally, the following equation shows the estimated relationship between the
natural gas consumption in the commercial/public and residential sectors.
(4.5)
Where and are coefficients and is the error.
3Due to lack of data, this study assumed that the relationship between the specific consumption of the
commercial/public and residential sectors remains constant over the period of analysis.
60
In the case of Argentina, using ordinary least squares (OLS) on historical data of
connected users by sector of (ENARGAS, 2015) based on
(4.2 the following expression was estimated:
(4.6)
(-6.71) (44.43) t statistic ()
Furthermore, to calculate the total energy consumption, the number of connected
users are multiplied by the average specific consumption per user obtained from
(ENARGAS, 2015) using Eq.4.3, which is estimated at 5.7 thousand cubic meters of
natural gas per year per user in the commercial and public sector.
In the case of Bolivia, by the lack of a long time series of number of users, Eq.
4.4 was applied based on the historical data for the period 1995-2012 (IEA, 2014),
obtaining the following expression.:
. (4.7)
(4.15) (15.74) t statistic ()
The same approach was applied to Chile using Eq. 4.4 with historical data for the
period 1980-2012 (IEA, 2014), since it has the same problem as Bolivia concerning
long time series of number of users:
(4.8)
(-4.13) (19.52) t statistic ()
Similar to the residential sector, the number of households was adopted as a
potential driver in order to distribute the additional demand among the regions in Chile
for the commercial and public sector.
4.1.3 Industrial Sector
To model energy consumption in the industrial sector some studies include fuel
prices as an input (HUNTINGTON, 2007; WANG and LIN, 2014), others suggest GDP
61
as a relevant variable to explain the trend component of the aggregated industrial
demand because it mainly summarizes both aggregated productivity of existing
industries and new industrial consumers (SÁNCHEZ-ÚBEDA and BERZOSA, 2007).
In addition, natural gas used by industrial consumers is, in general, not clearly
influenced by outdoor air temperature, and follows very different patterns (SÁNCHEZ-
ÚBEDA and BERZOSA, 2007).
Predicting fuel prices is a difficult task. Even when adopting the fuel projections
of other studies, the available historical data for fuel prices in the reviewed countries are
shorter than the minimum required performing a suitable correlation. In order to project
the natural gas consumption in the industrial sector until 2030, this study considered the
limitations of using a price-based model, and, alternatively, assessed the relationship
between the natural gas consumption in the industrial sector (IEA, 2014) and GDP in
real 1995 dollars at Purchasing Power Parities from (IMF, 2015).
The Johansen’s cointegration test allows the estimation of a long-run relationship
between non-stationary time series that become stationary at first differences. The test is
based on the following vector error correction model (VECM).
∑ (4.9)
where Yt is the data series vector; Π is the matrix containing the cointegration
relationships associated with the long-run dynamics and quantified by the error
correction term (ECT); represents the short-run dynamics (coefficient vector); and
the white noise.
This study uses only one explanatory variable (GDP) to predict the industrial
natural gas consumption (NGind) in order to avoid the problem of collinearity between
the explanatory variables. Annual data for the 1980-2012 period was transformed into
logarithm and the results for one lag4 showed cointegration between the natural gas
industrial sector consumption (NGind) and the gross domestic product (GDP) series.
This approach was used for Bolivia and Chile (See Annex Section A).
4According to both Akaike and Schwarz information criteria.
62
To project the natural gas consumption in the industrial sector of Chile, the effect
of the interruption of the Argentinian natural gas supply should be taken into the
consideration. From 2007 and 2009 the natural gas consumption fell dramatically in this
sector, substituted mainly by diesel and fuel oil, but with no significant changes in terms
of useful energy (see Annex A, Figure 10.1 and Figure 10.2). Therefore, in order to
project future natural gas consumption, we forecasted the useful energy in the industrial
sector using a VEC model. Once the useful energy was projected, in order to estimate
the natural gas consumption the share of natural gas attending the useful energy
increased to 50% in 20305 (Figure 10.1 in Annex A). This assumption is in accordance
with the diffusion of the natural gas consumption in the framework of cleaner fuel use
policies (MINENERGIA, 2015) and the impact of lower expected cost of natural gas
under a retake hypothesis of Argentinian supply (CORIA, 2009).
In the case of Argentina, the series were not correlated according to the Johansen
Test results, so a VEC model is not an appropriate approach. As an alternative, an
Autoregressive Integrated Moving Average Model, ARIMA (1,1,1), was applied.
In addition to the natural gas projections made using the VEC and ARIMA
approaches, major industrial projects that could make a significant change to the
predicted demand were identified; they were then added to the demand projected with
the econometric approach. This is because these projects lead to a discrete increase of
industrial natural gas consumption not devised by the econometric approach.
Regarding major industrial natural gas projects in Argentina, the petrochemical
company DOW has declared its intention to duplicate their current capacity in the
petrochemical complex of Bahia Blanca considering the developments in the Vaca
Muerta basin (REVISTAPETROQUIMICA, 2015). However, since neither official
numbers for capacities nor dates for the construction and commissioning of these
5This is a conservative value compared to other countries. For instance in Argentina, the
participation estimated for natural gas in the useful energy-excluding electricity- in the industrial sector in
2012 was 84%. In the case of Chile, the highest historical value for natural gas in the useful energy-
excluding electricity- was 24% in 2004.
63
projects have been published, we did not add other natural gas consumption in the
industrial sector besides those estimated by the autoregressive model.
In the case of Bolivia, YPFB built its first extraction plant for natural gas
liquids in Rio Grande, which can process up to 5 Mm3/d. This plant increased the local
consumption by 0.30 Mm3/d and produces up to 350 tonnes of LPG a day, allowing
Bolivia to become an LPG exporting country (YPFB, 2015a). Moreover, a second plant
is being constructed by YPFB near the Argentinian border6, in order to process the
contractual daily export capacity to Argentina of 27 Mm3/day. This plant will be able to
produce 2,250 tonnes of LPG per day and 1,650 tonnes of naphtha per day (Pagina
Siete, 2014). In addition to these two industrial plants, this study considered the
ammonia-urea plant at Bulo Bulo, Carrasco (Cochabamba), which will produce export-
oriented fertilizers. According to YPFB (2013), the natural gas consumption estimated
for this plant is 1.4 Mm3/day.
4.1.4 Transport Sector
Transport natural gas modelling was made using a bottom-up approach,
estimating vehicle fleet by type of vehicle and fuel used, and specific consumption of
energy by type of vehicle according to the fuel used (CHAVEZ-RODRÍGUEZ, M. et
al., 2014).
The vehicle fleet was modeled by applying the mathematical formulation of
DARGAY et al. (2007) as shown in (4.10).
( )
(4.10)
6 YPFB also plans to implement a petrochemical complex in this area and has already contracted an
international company to start the conceptual project (MHE, 2015).
64
where refers to vehicle ownership (vehicles per thousand people), is the
saturation level; denotes the population density, urbanization, and GDP the per-
capita income in real 1995 dollars evaluated at Purchasing Power Parities; α and β are
negative parameters defining the shape of the Gompertz function, t the year and the
random error. and are two dummy variables defined as:
{
(4.11)
This study assumed the same saturation level ( ) estimated in Dargay et al.
(2007) for the USA. For the adjustment parameters ( and ) and the parameters α, φ
and λ, the pooled time series coefficients obtained by DARGAY et al. (2007) were used.
GDP and population projections were obtained from IMF (2015) and UN-DESA
(2012), respectively. Just in the case of Argentina, to consider more optimistic values
than IMF (2015), we used the projections of PWC (2015) from 2017 onwards.
This compiled historical vehicle ownership from DNRPA (2015), INE (1999,
2004, 2014a) and INE (2013c) for Argentina, Bolivia and Chile respectively. Finally, β
was fitted minimizing both of the whole period and the last observed errors. Figure 4.3
shows the estimated projected vehicle ownership for Argentina, Bolivia and Chile.
Figure 4.3. Projected vehicle ownership until 2030.
0
100
200
300
400
500
600
19
80
19
82
19
84
19
86
19
88
19
90
19
92
19
94
19
96
19
98
20
00
20
02
20
04
20
06
20
08
20
10
20
12
20
14
20
16
20
18
20
20
20
22
20
24
20
26
20
28
20
30
vehicles/
1000people
Chile
Argentina
Bolivia
65
Data compiled in the research made by PULIAFITO et al. (2015) was used to
split vehicle fleet by types and by fuel for Argentina, CHAVEZ-RODRIGUEZ et al.
(2016a) for Bolivia, and, in the case of Chile, data from INE (2014a) was used. For the
sake of simplicity and due to the lack of historical data, the share of each type of vehicle
was kept constant in the whole period assessed for Argentina and Chile, and for Bolivia
the projections made by CHAVEZ-RODRIGUEZ et al. (2016a) were adopted.
In the case of Argentina, the share of natural gas by type of vehicle in 2012
was assumed to increase by 50% until 2030 in the baseline scenario7. For Bolivia, the
cylinder of natural gas (CNG) conversion statistics of EEC-GNV (2013) and 2012 data
reported by INE (2014c) were used to obtain the share of fuel by type of vehicle. Then,
it was assumed that the share of natural gas would increase by 50% until 2030.
Finally, in Chile, the conversion of private vehicles to natural gas is currently
only permitted for taxis and commercial purposes, as a consequence of a policy to avoid
losses of fiscal incomes of the oil products’ sales (LA TERCERA, 2013). For scenario
purposes, however, we considered in our projections that Chile will reach a share of
natural gas in light vehicles similar to the current fleet of Argentina (21%) and in public
transportation similar to the current fleet of Bolivia (31% for minibuses and 2% for
buses).
7This is still conservative for the penetration of natural gas in the transport sector. As an upper
limit, for instance, in Bangladesh natural gas vehicles had 61% of market share in 2010 (NIJBOER,
2010).
66
5 Natural Gas Supply Technologies in TIMES-
ConoSur
TIMES-ConoSur represents the natural gas supply of the energy system of the
Southern Cone covering the following components: primary energy supply (extraction,
losses and imports/exports), transformation (including the processing of natural gas and
the separation of liquids of natural gas and also the transformation of natural gas into
electricity) and final consumption among the end-use sectors (residential, commercial
and public, industry and transport). As previously mentioned, the demand sectors are
modelled in an aggregated simplified way and calculated exogenously, then the final
values are inserted in TIMES-ConoSur with their respective profile curves.
The main model outputs are investment decisions regarding new domestic
production capacity and flows of natural gas allocated across the 6 regions, the installed
capacity and operation of the different natural gas technologies (NGPP, LNG Plants,
Gas to Liquid Plants, etc.); the operation of the power sector and the natural gas
consumed in this operation; the energy trade between the 6 regions modelled and with
external markets; and the overall system costs, along with CAPEX and OPEX by
technology and regions.
The model is structured in four main components: upstream extraction of natural
gas (divided by non-associated conventional gas, non-associated unconventional gas
and associated gas), midstream (processing, transport and trade of natural gas), power
generation and final demand. Figure 5.1 shows the structure of TIMES-ConoSur.
Natural gas and power can be traded across the six regions, and import and
export of LNG from and to the rest of the world are also possible. Therefore, TIMES-
ConoSur can model the natural gas market in the six regions in an integrated manner.
The following sections describe how the power generation sector, and the upstream and
midstream for natural gas were modelled.
67
Figure 5.1. TIMES-ConoSur architecture
68
Table 5.1. Acronyms description used in Figure 5.1 for process technologies
69
5.1 Power Generation Modelling
There are mid-term power expansion studies for the four countries assessed
(AGEERA, 2012; CNE, 2015; EPE, 2015b; MHE, 2012). We input the power
expansion capacities by technologies and the electricity demand projected in these
studies8 in TIMES-ConoSur (See Figure 5.2). There are two electricity demands: one is
the average annual demand, and the other is the maximum annual demand, which is
calculated based on the historical relationship between average and peak demand. As
shown in Figure 5.2, in most of the modelled countries the capacity expansion will
strongly rely on hydropower and non-conventional renewable energies such as wind and
solar. It is important to highlight that this figure shows the installed capacity, instead of
the electricity produced by these renewable technologies to balance the supply and
demand of electricity (Figure 5.4,Figure 5.5 and Figure 5.6).
As for Bolivia, the capacity expansion plan had taken as reference a time horizon
up to 2022, from 2023-2030 we used the extrapolated expansion of CHAVEZ-
RODRIGUEZ et al. (2016a) and only for this country the decision to expand the
capacity of natural gas power plants is endogenous in the model.
As the electricity demand is projected on an annual basis, we have kept the same
commodity fraction9 shape of the base year (Figure 5.3) for future years. With these
power-expansion capacities and a fixed electricity demand, the model carries out a
least-cost operation of the power system, which has the natural gas consumption as one
of its outcomes. For this operation, we incorporated neither ramp-up, ramp-down nor
down time constraints.
Each power generation technology used in these countries was modelled based on
general technical and economic characteristics such as capacity, conversion efficiency,
8 The idea here is not to validate the mentioned studies, but to use them to test the developed
model.
9 Commodity fraction is referred as the distribution of the yearly total energy in the defined time
slices. The sum of the commodity fractions across the year is 1.
70
and costs. Costs for conventional technologies were based on BLACK & VEATCH
(2012) and those for renewable technologies were based on IRENA (2015). These
parameters for power generation technologies are shown in Table 10.6.
The availability of renewable resources, such as hydro, wind and solar, were
incorporated based on historical capacity factors from the literature. Capacity factors of
hydro and wind were modelled on a monthly basis (Figure 5.4 and Figure 5.5) and in
the case of solar on an hourly basis (Figure 5.6). Finally, fuel prices for power
generation were based on the current costs in each country. This is a relevant feature
since in some countries oil products for power generation are subsidized(CHAVEZ-
RODRIGUEZ et al., 2016a).
71
Figure 5.2. Power Generation Capacity Expansion by technologies.
Source: AGEERA (2012) CHAVEZ-RODRÍGUEZ et al. (2016); CNE (2015);
EPE (2015b); MHE (2012); SORIA (2016).
72
Figure 5.3. Commodity Fraction of the electricity demand modelled in the base year.
Source: CAMMESA (2015); CNDC (2016); CNE (2016b); ONS and SORIA (2014)
Figure 5.4.Monthly capacity factors adopted for Hydropower plants by countries.
Source: CNDC (2016); CNE (2016b); ONS (2016); SECRETARIA DE
ENERGIA (2014)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Capacity Factor
Bolivia
Argentina
Chile
IntegratedBrazil
North Brazil
73
Figure 5.5. Monthly capacity factors adopted for wind power plants by countries.
Source: CNDC (2016); CNE (2016b); SECRETARIA DE ENERGIA (2014);
SORIA (2016)
Figure 5.6. Hourly capacity factors adopted for solar power plants by countries
(PV solar power). Source: Based on CNDC (2016); CNE (2016b); SECRETARIA DE
ENERGIA (2014); SORIA (2016)
0
0.1
0.2
0.3
0.4
0.5
0.6
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Capacity Factor
Bolivia
Argentina
Chile
Brazil
00.10.20.30.40.50.60.70.80.9
1
M0
1h
01
M0
1h
07
M0
1h
13
M0
1h
19
M0
2h
01
M0
2h
07
M0
2h
13
M0
2h
19
M0
3h
01
M0
3h
07
M0
3h
13
M0
3h
19
M0
4h
01
M0
4h
07
M0
4h
13
M0
4h
19
M0
5h
01
M0
5h
07
M0
5h
13
M0
5h
19
M0
6h
01
M0
6h
07
M0
6h
13
M0
6h
19
M0
7h
01
M0
7h
07
M0
7h
13
M0
7h
19
M0
8h
01
M0
8h
07
M0
8h
13
M0
8h
19
M0
9h
01
M0
9h
07
M0
9h
13
M0
9h
19
M1
0h
01
M1
0h
07
M1
0h
13
M1
0h
19
M1
1h
01
M1
1h
07
M1
1h
13
M1
1h
19
M1
2h
01
M1
2h
07
M1
2h
13
M1
2h
19
Capacity Factor
Bolivia Argentina Chile Brazil
74
Figure 5.7. Fuel Prices assumed for power generation. Source: Based on ANH
(2015); ANP (2015); CAMMESA (2015); CNDC (2013); CNE (2015); EPE (2016a).
The national electricity trade between the regions is relevant for the power
generation modelling. The expansion of the interconnection capacity is shown in Table
5.2 for the two countries modelled by two regions in TIMES-ConoSur: Brazil and
Chile. In the case of Brazil, the expansion of the interconnections between the North
Region and the other regions of the SIN (Interconnected National System) projected by
EPE (2015) up to 2024 were maintained until 2030 (see Table 5.2). In the case of Chile,
the most important interconnected systems, the “Sistema Interconectado Central” and
the “Sistema Interconectado Grande Norte” are planned to be interconnected in July
2017 through a 600-kmtransmission line of 500 kV and 1500 MW(E-CL, 2015); in
order to be conservative, the start of this interconnection was set to the end of 2017.
0
5
10
15
20
25
MUS$/PJ
Argentina
Bolivia
Brazil
North Chile
Central-South Chile
Diesel Fuel Oil Coal
75
Table 5.2. Interconnection capacity for domestic energy trade considered between
regions in Brazil and Chile. Source: Based on E-CL (2015); EPE (2015b)
Year North Brazil - Integrated Brazil (MW)
North Chile - Central and South Chile (MW)
2012 14461 0
2013 14461 0
2014 14461 0
2015 14461 0
2016 20271 0
2017 20953 0
2018 26680 1500
2019 30680 1500
2020 39085 1500
2021 40985 1500
2022 42985 1500
2023 44985 1500
2024 48285 1500
2025 48285 1500
2026 48285 1500
2027 48285 1500
2028 48285 1500
2029 48285 1500
2030 48285 1500
Regarding the international connections, Brazil and Argentina have two
interconnections, one of 50 MW and 2200 MW, with converters of frequency 50/60 Hz
with systems back-to-back. Historically, the energy trade using these interconnections
only occurred due to contingency circumstances.
In the case of Chile and Argentina, there is an international transmission line of
640 MW (CEARE-UBA, 2004). From January until June of 2016, Chile exported
105GWhto Argentina (CAMMESA, 2016; LA TERCERA, 2016); this was the first
electricity export from Chile since 2009.
Bolivia is planning the construction of an 110-km transmission line of 550 kV
to export electricity to Argentina. Regardless of the fact that the financing from the
Central Bank of Bolivia for this project has been approved (ENDE, 2016), this
international interconnection was not incorporated in the model, as any public data
about the capacity is available at the moment of the development of this thesis.
76
International electricity trade using existing transmission lines was set free in the model
to run under an optimized operation from 2016 onwards.
5.2 Upstream Modelling
An upstream production facility involves wells, platforms, storage, piping and
separation facilities used in the production, extraction, recovery, lifting, stabilization,
separation or treating of the hydrocarbon produced. After the separation process, there
are three main products separated from free-water and solids: crude oil, wet gas, and
condensates.
Wet gas is natural gas that contains methane (typically less than 85% methane on
a molar basis) and ethane and other longer chain hydrocarbons; these last are the so-
called natural gas liquids (NGLs) and plant condensates that will be separated further in
a Natural Gas Processing Plant. Natural Gas Liquids involves a group of hydrocarbons
including ethane, propane, normal butane, isobutane, and natural gasoline (EIA, 2016b).
Table 5.3. Natural gas classification according to gas-to-oil volumetric ratio (v/v)
Classification Gas-to-Oil Ratio (v/v)
Gas non-associated >10 000
Gas Condensate >5.000
Gas dissolved in oil >1 000
Gas separated from oil <1 000
Source: THOMAS (2004)
Wet gas can be found as associated or non-associated gas (see Table 5.3).
Associated dissolved gas is produced in oil fields where natural gas is found either as
free gas (associated), or gas in solution with crude oil (dissolved)(EIA, 2016b). Non-
associated gas is produced by gas wells10
in a gas condensate fields or in gas fields. Gas
condensate produced in gas condensate fields consist predominantly of methane (C1)
10 According to the RRC (2003) a as well is defined as “any well which produces more than
100,000 cubic feet of natural gas to each barrel of crude oil during the same producing horizon”.
77
and other short-chain hydrocarbons, but also contains long chain hydrocarbons. An
average gas condensate usually contains 70–75 mol% methane and 5–10 mol% C7+
fraction, with the rest distributed between the non-hydrocarbons and the intermediates
(DANDEKAR, 2015). Gas fields produce a more “dry” gas in which liquids fractions
are less significant. For modelling purposes, we consider a gas condensate commodity
for non-associated gas wells of both gas fields and gas condensate fields. Figure 5.8
shows the proportion of condensates and wet gas on an energy basis by each country
adopted for modelling upstream in TIMES-ConoSur.
Figure 5.8. Condensates to Wet Gas ratio on energy basis assumed for non-
associated gas for each region. Source: SGI (2016)
It is worth to discuss the definition of condensates since this is a relevant by-
product for the economics of natural gas and it is commonly neglected in energy
planning models11
.
According to the Code of Federal Regulations of USA (CFR, 2000) condensates
are defined as: “hydrocarbon liquid separated from natural gas that condenses due to
changes in the temperature or pressure, or both, and remains liquid at standard
conditions”. However, this definition encompasses the light liquid hydrocarbons
recovered from natural gas fractioning processing plants (“plant condensate”) and those
11For instance, from the natural gas models reviewed in Figure 5.8, only the NEMS model was
identified that incorporates condensates.
0.000.020.040.060.080.100.120.140.160.18
boe Condensates/ boe Wet Gas
Bolivia Argentina Chile Integrated
Brazil North Brazil
78
separated from field separators in the production facilities (“lease condensate” or “field
condensate”) from associated and non-associated natural gas wells(EIA, 2016b).
There is not a consensus about the definition of condensates based on API gravity
(KEMP, J., 2014). However, it is often referred to those liquids hydrocarbons with an
API gravity of more than 50 degrees (IEA, 2015a).In terms of the chemical composition
of condensates, the lightest molecules consist of five carbon atoms and progress to
heavier ones. Condensates are often referred to as C5+.
The chemical composition mix varies between lease condensates and plant
condensates. Since plant condensate is the product of a processing plant, its
specifications are in tighter ranges than lease condensates, predominating pentanes (C5)
and hexanes (C6) and very small quantities of C7+ (RBN ENERGY, 2012). That is why
in some countries such as Bolivia, Brazil and USA, plant condensates are called also
“natural gasoline”. Condensates are used in markets as diluent for heavy crudes, the
petrochemical feedstock for steam crackers, refinery and blending feedstock for further
transformation, and even boiler feed in final use (EPRINC, 2015). The condensates
production is incorporated in the model as an output of the non-associated gas wells,
and this has a significant impact on the shadow price of the natural gas produced given
the opportunity cost of condensates. The price for condensates produced was assumed
to be 44US$/bbl, based on PLATTS (2015).
In an aggregated way, to incorporate the economics and rationale of gas
production, natural gas production was divided into three categories: Associated Natural
Gas (includes conventional and unconventional resources), Conventional Non-
Associated Natural Gas and Unconventional Non-Associated Natural Gas.Before
explaining the modelling of natural gas production, the consumption and losses in the
upstream are discussed as this affects both non-associated and associated gas net
production.
79
5.2.1 Upstream natural gas consumption and losses
Natural gas consumption and losses in the upstream can be divided into four
categories: self-consumption, flaring, venting and fugitive emissions.
Based on IOGP (2015), natural gas consumed in lease fuel and operations is
related to natural gas re-injection for oil recovery, powering compressors and pumps,
power generation, heating and steam production among other activities. Flaring is the
controlled burning of hydrocarbons produced in the course of petroleum exploration and
production operations. Venting is the controlled release of unburned gas to the
atmosphere. The fugitive emissions are unintended emissions released into the air, other
than those from stacks or vents from the processing, transmission, and/or transportation
of fossil fuels; this study does not considerthese losses as they are not reported. Table
5.4 shows the consumption and losses used for modelling based on historical values.
Table 5.4. Consumption and Losses parameters considered for upstream
modelling
Category Argentina1 Chile2 Bolivia3 Brazil-GIN4 Brazil-NR4
Gross gas produced 100% 100% 100% 100% 100%
Self-Consumption (Energy) 12.0% 12.0%* 1.5% 13.1% 5.4%
Losses (flaring+venting) 2.3% 4.3% 0.9% 5.1% 2.6%
Net gas produced (without re-injection 85.7% 83.6% 97.6% 81.8% 92.0%
Re-injection 0.2% 0.0% 0.0% 18.0% 54.3%
Net gas produced (including re-injection) 85.5% 83.6% 97.6% 63.8% 37.7%
*Estimated.
Source: Based on SECRETARIA DE ENERGIA (2015b)1,EIA (2015)
2,
YPFB(2015b)3, ANP (2015)
4
Table 5.4 shows how heterogeneous is the management of consumption and
losses in the upstream in the countries analysed. Their consumption and losses can be
explained by factors such as gas-to-oil ratio, reservoir characteristics, location and
logistics, the age of the fields, use of hydrocarbon recovery techniques, regulations and
contractual issues (IOGP, 2015). For instance, flaring is not only explained by safety
80
and technical reasons but also by the lack of export infrastructure, which could happen
when hydrocarbon fields are far from consumption centres or gas pipelines in countries
with no penalties for flaring. However in the case of the Brazilian pre-salt, where flaring
and venting is controlled, the high CO2content in the natural gas produced, estimated in
a range of 10% and 45% mol (SILVA, 2015), and the costs and topside footprint
required to separate the CO2, explain why a fraction of the produced gas is being re-
injected in the fields. For energy consumption, mature or remote fields usually consume
more energy than other fields (IOGP, 2015). In addition, energy consumption is also
significant for compression and re-injection of CO2 in the reservoirs, as it would be in
pre-salt fields (SILVA, 2015).
5.2.2 Non-Associated Natural Gas Production Modelling
There are two basic approaches to forecasting the production of non-associated
natural gas: top-down and bottom-up (BRANDT, 2010; REYNOLDS and BAEK,
2012). The top-down approach requires less data than the bottom-up one (CHAVEZ-
RODRIGUEZ et al., 2015). There are stylized top-down methodologies, such as
Hubbert models (HUBBERT, M. K., 1956), Generalized Weng model (CHEN, Y.Q.,
1996), and HCZ model (CHEN, Y.; HAO, MINGQIANG, 2013). Top-down approaches
might be convenient for a global scale natural gas forecast, but not for a country with a
reduced number of wells drilled as explained by CHAVEZ-RODRIGUEZ et al.
(2016a).
A bottom-up methodology requires extensive and disaggregated data, such as
reserves and resources and current production by field, production costs, government
take, natural gas selling price, discount tax, etc.
In the countries of the Southern Cone, data regarding upstream is limited. Only
Argentina has public disaggregated data by fields. Therefore, given this data availability
constraint, a hybrid approach was elaborated and adopted, based on the sum of
production curves of a typical conventional natural gas field profile and an aggregate
production profile defined by the total natural gas consumption and exports of each
country, plus losses. The method is based on
81
(5.1,
and follows the procedure described subsequently.
[ ] ⁄ ∑
∑
∑
∑
∑ (
)
∑ ∑
∑ ∑
∑ ∑
∑ ∑
{
(5.1)
where is the natural gas demand projected for year"t", is the loss production factor
(flaring,venting and self-consumption), is the base-year, is the first-year of
production of the fields, is the number of fields that began production in year,
are the reserves/resources quantified in , is the production function of the
“typical non-associated gas field” assumed for a conventional natural gas field, x,y, w
82
are the last year for expanding fields producing proven, probable and possible reserves
(they can be read also as the first year of fields producing probable, possible and other
resources respectively). T is the lifespan assumed for the field (18 years for this study),
is the production of the base-year considered as the current capacity
production of developed proven reserves, is the initial decline rate, b is a constant
commonly known as the “Arps” factor.
The method is carried out according to the following steps:
Step 1: Estimating total natural gas production requirement based on the
natural gas total demand and exports plus a loss factor (flaring, venting and
self-consumption).
Step 2: Defining a decline curve for the existing natural gas production
capacity. For that, the current production of developed reserves was
modelled using an ARPS (1945)’s hyperbolic decline curve (Eq. 5.2).
Eq. 5.2
[ ] ⁄
where t is time(years), qi is the initial surface rate of flow at t=0, Di is the initial
decline rate, b is a constant commonly known as the “Arps” factor. As commented
by ADEBOYE et al. (2011), even in a well that follows an exponential decline
solution, the total decline curve analysis production from the reservoir or field would
be better estimated using hyperbolic decline model. Values for of 15% and a b of
0.10 were used based on FERIOLI (2013)’s curve production for a typical
conventional natural gas well of Argentina.
83
Figure 5.9. Natural gas production profile of a typical field.
Source: Based on CAMAROTA (2007).
Step 3: For the simulated years, the gap between estimated natural gas
production requirements and the base year natural gas production capacity
curve is completed by new fields. It is assumed that these fields will
produce, throughout their lifespan, available resources sequentially from
proven remaining reserves to a fraction of contingent resources. This
procedure is repeated for each simulated year, limited to an eventual
situation of exhaustion of the ultimate recoverable resources. The following
figure illustrates the results of this procedure, which was implemented by
CHAVEZ-RODRIGUEZ et al. (2016a).
84
Figure 5.10. Estimated natural gas production requirement and natural gas
production capacity curves according to reserves/resources classification, an example of
Bolivia. Source: CHAVEZ-RODRIGUEZ et al. (2016a)
As stated by CHAVEZ-RODRIGUEZ et al. (2016a), this method allows
dealing with the lack of disaggregated data. It provides useful insights into upstream
annual investment requirements and temporal depletion of currently accounted reserves
and recoverable resources. Moreover, it accounts for the reserves already committed to
an installed production capacity, hence indicating the need to develop new
reserves/resources over time.
Using production data for oil and gas production by fields in Argentina
(SECRETARIA DE ENERGIA, 2015b) and values showed in Table 5.3, non-associated
natural gas fields were filtered and modelled using this approach. This involved
conventional and unconventional resources.
For Bolivia and Chile, based on local knowledge, almost all natural gas
produced is sourced in gas or gas-condensate fields. Then, for the sake of simplicity, for
current capacity of natural gas production and reserves and resources,the proposed
approach of non-associated natural gas is used.
85
Finally, in Brazil, non-associated natural gas such as the Mexilhão in the Santos
Basin, Manati in the Camamu Basin and new fields in the Solimões Basin were
modelled using this non-associated gas modelling methodology.
5.2.3 Associated Natural Gas Production Modelling
Associated natural gas production relies on the dynamics of crude oil production,
which is assumed to not having either logistics constraints or dependence on the
domestic market. Oil production also has a shorter length between discovery and
production start, differently from non-associated natural gas projects (SÄLLH et al.,
2015). For instance one of the largest gas fields in South America, Camisea, was
discovered in 1985 but only after almost 20 years the field was developed and the
production started. Furthermore, associated natural gas is considered as a “by-product”
and, in the past, it was a common practice to flare it, since the project is already paid by
the liquid productions and additional investments are required to monetize the
associated gas (OGJ, 2002).
Consequently, in order to forecast the associated natural gas production,oil
production models were developed. For that purpose, we used a Multi-Hubbert
approach (CHAVEZ-RODRIGUEZ et al., 2015; LAHERRERE, 1997) mixed with the
variant of MAGGIO and CACCIOLA (2009) using different scenarios of EUR
(Estimated Ultimate Recovery). The following equations summarize the approach.
N
i iMi
iM
ttbk
PP
1 ).(cosh1
.2 (5.3)
2
2
1
)ln()11ln(4
k
kk
b
PEUR Mi
i
(5.4)
86
N
i
iEUREUR1
(5.5)
where EUR is the estimated ultimate recovery (or ultimate recoverable reserve), Q is oil
production at time t, PM peak production, tM the time of the peak, b is a parameter which
accounts for the slope of the curve, i represent each cycle, and N the total number of
cycles modelled. The constant k (0 < ki ≤ 1), introduced by MAGGIO and CACCIOLA
(2009) aims at the improvement of the fitting.
Oil production was modelled only for Argentina and Brazil, where the production
of associated natural gas is relevant. In the case of Bolivia and Chile it was assumed that
all the gas produced is non-associated.
The oil production projection for Argentina was based on historical oil production
obtained from IAPG (2015). The EUR is composed of 3P reserves and Contingent
Resources were used. In the case of Argentina, a k value of “1” led to a satisfactory
fitting (Figure 5.11). Therefore, the basic Multi-Hubbert with no variant was suited for
this country.
87
Figure 5.11. Multi-Hubbert Curve for oil production calculated for Argentina.
Source: Based on IAPG (2015)
As shown in Figure 5.11, there are three Hubbert cycles that explain the oil
production in Argentina. The first one is related to the openness to private oil and gas
companies during the 50s and 60s and the developments in the Golfo San Jorge Basin
(ECONLINK, 2008; GADANO, 1998); the second cycle is explained by the new cycle
of production in the second half of the 80s and the 90s, promoted also by large amount
of concessions to private companies and the market-oriented precification reforms for
the petroleum produced domestically (VÁSQUEZ, 2016). Finally, the third cycle
reflects the production of the remaining EUR, the total EUR discounted by the
production of the first and second cycle.
To project Brazilian offshore oil production, this study updated the work of
SARAIVA et al. (2014) with production data for the 2013-2015 period. This updating is
relevant since, in the last years, Post-Salt production has declined steadily. With this
declined production in the last years, it was necessary to add another cycle to increase
the Post-Salt production in the future. In addition, there are more observations about the
Pre-salt’s production (MME, 2016a) that has helped to better define the b parameter,
with a faster pace of production than the one adopted by SARAIVA et al. (2014). The b
and k parameters for Post-Salt were also updated (See Table A.1).
0
10
20
30
40
50
601
91
1
19
15
19
19
19
23
1927
19
31
19
35
19
39
19
43
19
47
19
51
19
55
19
59
19
63
19
67
19
71
19
75
19
79
19
83
19
87
19
91
19
95
19
99
20
03
20
07
20
11
20
15
2019
20
23
20
27
MMm3/year
H1 H2 H3 H1+H2+H3 Historical Production
88
Once the oil production curves were obtained, the natural gas-to-oil ratios were
used to project associated natural gas production. For Argentina, a gas-to-oil ratio of 0.3
Mm3 of natural gas per Mm3 of crude oil (0.036 Mm3/Mbbl) was used, based on
historical production (SECRETARIA DE ENERGIA, 2015b). In the case of Brazil, the
same procedure was adopted based on historical natural gas-to-oil ratio production for
Post-Salt and Pre-Salt (Figure 5.12). Pre-Salt oil fields usually have a higher gas-to-oil
ratio than Post-Salt fields, but as happened at the Lula field, 35% of this gas is CO2 (DE
MORAES CRUZ et al., 2016). Therefore,in terms of natural gas (or methane plus C2
and C3+)-to-oil ratio, Post-Salt fields were modelled with slightly higher ratios than
Pre-Salt in most of the past years.
Figure 5.12. Historical natural gas to oil ratio from oil fields in Post-Salt and Pre-
Salt in Brazil. Based on MME (2016a)
Natural gas production in North of Brazil, specifically in the Amazonas State, is
mainly associated with oil production. Therefore it was necessary to project oil
production in this region. For this purpose, an ARPS (1945)’s hyperbolic equation was
used. The “ ” factor was calculated using the decline rate production of 2013-2014
(9.3%) (ANP, 2015). Furthermore, the “b” (Arps factor) was estimated aiming that the
accumulated oil production up to 2030 wouldbe equal to the Total Oil Reserves (3P)
reported in 2014 for the Amazonas State in ANP (2015) (89.6 MMbbl) (Figure 5.13).
0.0000
0.0050
0.0100
0.0150
0.0200
0.0250
0.0300
2011 2012 2013 2014 2015
MMm3/Mbbl
Pre-Salt
Post-Salt
89
Figure 5.13. Oil production in the North Region projected to 2030. Based on:
ANP (2015)
In spite of the declining oil production, associated natural gas in the Amazonas
State is growing; this means that the gas-to-oil ratio has increased over time. Moreover,
the proportion of the gas re-injected has declined over the last years, especially since the
start of operations of the Urucu-Coari-Manaus gas pipeline. Taking into account these
factors, the ratio between and
oil produced was projected using the data between 2010-2014 (Figure 10.3 in Annex
A).
Following this approach, the associated natural gas production accumulated from
2015 up to 2030 would sum 23.7 Mm3. The difference between this accumulated
projected value of associated natural gas production and the total natural gas reserves
(3P) reported by ANP (2015) for 2014 (74.5 Mm3) will be considered as non-associated
natural gas total reserves (50.8 Mm3) and modelled using the bottom-up approach of
CHAVEZ-RODRIGUEZ et al. (2016a).
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
19
99
20
01
20
03
20
05
20
07
20
09
20
11
20
13
20
15
20
17
20
19
20
21
20
23
20
25
20
27
20
29
MMbbl
Observed Projected
⌊ ⌋ ⁄
= 9.3% b =0.45
90
5.2.4 Associated Natural Gas Options
Considering legislation constraining flaring (ANP, 2000), the net gas produced
after losses and energy consumption can either be re-injected in the oil fields or
monetized.
Re-injection of gas is a good alternative for the purpose of improving oil recovery
in areas lacking natural gas processing or transport capacity, or simply given the more
attractiveness of oil compared to gas. Furthermore, the natural gas can perhaps be
produced later if the production capacity becomes available (HINDERAKER and
NJAA, 2010).
In the case of the Post-Salt in Brazil, reinjection of associated natural gas has been
practised as a secondary recovery method at the Campos Basin. In the case of Pre-Salt
associated gas requires a separation process to remove the high content of CO2. The
lack of space at the Floating Production, Storage and Offloading (FPSO) unit to purify
all the associated gas produced (SILVA, 2015) is the major reason behind the
reinjection of natural gas considering the flaring constraints.This reinjection is
considered in the rates shown in Table 5.4. Somehow, this re-injected gas could be
considered as contingent resources that in the future may become commercially
feasible.
According to SILVA (2015), the CO2 separation by membranes is preferred over
the chemical absorption of CO2 due its lower footprints and the limitation of space in
the FPSO units. Consequently, based on SILVA (2015), the capital costs estimated
considering the skid of membranes and compressors resulted in 29 MUS$/Mm3 of
natural gas produced per day of capacity. Figure 5.14 shows in detail the process of
separation and treatment of the natural gas in a Pre-Salt FPSO unit.
91
Figure 5.14. Natural Gas Separation and Processing in the FPSO. Source: DE
MORAES CRUZ et al. (2016)
To include these aspects in the model, TIMES-ConoSur considers a reinjection
technology for natural gas additional to the values shown in Table 5.4, as an option for
lack of market or infrastructure for the associated natural gas produced. Monetization
options considered in the TIMES-ConoSur model for Brazilian associated gas were:
Pipelining it to the coast, Liquefaction through a Floating Liquefaction Natural Gas Unit
(FLNG), and investing in an Offshore GTL (gas-to-liquids) plant. To implement any of
these monetization options in the model, CAPEX investment relatedto a process for
removing the high content of CO2 is performed in the model as well based on costs
reported by SILVA (2015).
Transport of the gas from the fields to the main pipeline network may have
significant costs, especially when the producing platforms are in remote locations, such
as in offshore Post-Salt and Pre-Salt oil fields. For instance, Lula and Libra Pre-Salt oil
fields are located about 300 km and 170 km offshore the coast, respectively, whereas
Marlim and Roncador Post-Salt oil fields are 110 km and 125 km offshore the coast,
respectively. Due to longer distances, the average transport costs for Pre-Salt (estimated
at 50 MUS$ per Mm3/day of capacity) were higher than for Post-Salt (estimated at 36
92
MUS$ per Mm3/day of capacity). It is worth to mention that the capital costs for
upstream transport in non-associated gas fields are already included in the CAPEX NPV
per boe values shown inTable 5.6.
In the case of the GTL plant, taking into account the limitation of space and
weight on the FPSO, the low-temperature Fischer-Tropsch technology evaluated in
CASTELO BRANCO et al. (2010) was considered for associated natural gas from
offshore fields in Brazil - see process flowchart in Figure 5.15. This technology
produces a long chain paraffinic crude oil and sells it at a price of 60US$/bbl, which is
an assumption of this study.
Figure 5.15. Low-Temperature Fischer-Tropsch Process Overview. Source: SBM
OFFSHORE (2014)
For the GTL plant, the consumption of 1780 m3 of NG (this volume of NG
contains an equivalent of 74,005 MJ) to produce 1 m3 of syncrude (this volume of
syncrude contains an equivalent of 37,325 MJ)was considered. The CAPEX of 100 000
US$/barrel of capacity per day and OPEX of 6 US$/barrel processed used by
CASTELO BRANCO et al. (2010) was adopted.
5.2.5 Reserves, Resources and Costs
To apply the methodology of CHAVEZ-RODRIGUEZ et al. (2016a) for non-
associated natural gas described in the previous sections, the reserves and resources of
non-associated gas are required.
93
Reserves and resources statistics usually does not detail non-associated natural gas
and associated natural gas (ANP, 2015, 2016b; IEA, 2015b; SECRETARIA DE
ENERGIA, 2015c; USGS, 2016; YPFB, 2014). Where data for reserves and natural gas
and oil production is available at the field level, as in Argentina, by using the ratios
presented in Table 5.3 it is possible to infer which fields are oil or gas fields. However,
in the other countries, assumptions had to be made based on aggregated data and the
knowledge of the local hydrocarbons production – e.g, in Chile and Bolivia natural gas
domestic production relies only on gas fields.
As the methodology of CHAVEZ-RODRIGUEZ et al. (2016a) does not
incorporate a process to increase reserves over time through exploration of
undiscovered resources or reserves growth mechanism based on contingent resources,
these resources are incorporated according to a static EUR definition. The EUR
definition is fundamental to elaborate long-term supply projections. For this purpose,
many modelers have used 2P reserves for its either conservative scenarios or “best
estimates” scenarios (CHAVEZ-RODRIGUEZ et al., 2015; MCGLADE, C.; EKINS,
2015; OWEN et al., 2010; SARAIVA et al., 2014; SZKLO et al. 2008).
The aim of this study is not an accurate forecast of natural gas production, but to
assess the required resources and investments efforts in upstream. Based on this,
additional categories of reserves and resources, such as Possible Reserves, Contingent
Resources and Yet-to-find Resources, can be included besides of the 2P reserves to
define the EUR.
It is worth to say that by definition the Possible Reserves are inside the boundary
of economically recoverable resources (ERR); however, they present uncertainties
related to technical parameters such as Original Oil/Gas-in-place, recovery factor, etc.
According to SPE (2007,pp.29), the Possible Reserves classification “may be assigned
to areas of a reservoir adjacent to Probable where data control and interpretations of
available data are progressively less certain”. One can expect that through increasing
information with exploration and development this “less certainty” would gradually
reduce.
94
Contingent Resources classification is only assigned to discovered accumulations.
Therefore, the risk component related to non-discovery is null, or the risk of
commerciality of these resources (becoming reserves) depends only on the chance of
development (SPE, 2011). In other words, a discovered resource classified as
Contingent Resource can be reclassified as Reserve once the contingency that does not
allow the project to turn it commercial is overcome(SPE, 2011). According to SPE
(2011) this contingency could be even non-technical, for instance not getting an
environmental approval, or technical issues– e.g.the quantities considered to be too
small to exploit. Moreover, the expansion of the transport and upstream infrastructure to
reach the area of the contingent resource could contribute to turn this resource into
reserves. In addition, for a resource to become reserve a reasonable timetable for
development is required, usually five years is adopted as a benchmark (SPE, 2007).
However, for a long-term production projection, this five-year period would not be a
limitation.
Another factor that could make contingent resources feasible is the development
of a market to at least absorb the minimum expected production that economically
justifies the development of these resources(SPE, 2007). Consequently, in the long-
term when originally estimated reserves are depleted and the domestic consumption and
exports are increasing, it is expected a market gap to absorb part of the contingent
resources and re-classify them to reserves.
To assess Yet-to-find resources , that can apply for undiscovered resources or
prospective resources (as it is used by the SPE), the methodology of (SPE, 2011)
considers the chance of discovery and a chance of development. In the chance of
discovery, there are estimated quantities of petroleum that would be recoverable (SPE,
2011). Depending on the methodology to assess these resources the quantities used will
vary significantly (MCGLADE, C. E., 2013). The most comprehensive and updated
estimates of undiscovered resources is the study performed by USGS in 2012(USGS,
2012).In this study, a fraction of contingent and yet-to-find resources were aggregated
in the “Other Resources” category included in the EUR used for each country (Table
5.5).
95
Therefore, to account Argentinian EUR this study used the 3P reserves of natural
gas reported by SECRETARIA DE ENERGIA (2015c). Other resources were
composed by Contingent Resources reported by SECRETARIA DE ENERGIA (2015c)
and mean estimates of undiscovered gas of USGS (2013).
For Bolivia, the 3P reserves of natural gas reported by YPFB (2014) and the mean
estimates of undiscovered gas of USGS (2013) were considered.
In Chile, there is no official open publication of a governmental entity about
natural gas reserves. Cedigaz and the Oil and Gas Journal have reported proven reserves
of 41 and 98 billion cubic meters by the end of 2013 (IEA, 2015b). To be conservative,
we have used the values of Cedigaz as the proven reserves and the difference to the Oil
and Gas Journal values as probable reserves. Other resources include mean estimates of
undiscovered gas of USGS (2013). According to a USGS (2016) assessment in Zona
Glauconita in the Magallanes Basin there are 2.46 TCF (95% probability) of tight gas
resources.These unconventional resources were also included into the EUR.
In the case of Brazil, there are no statistics disaggregating the natural gas reserves
as associated or non-associated. ANP just publish proven reserves(1P) and total reserves
(3P) by location (onshore or offshore) and by States (ANP, 2015). Recently, ANP
published contingent resources and the reserves by basins (ANP, 2016b). Having this
limitation of data in mind, the non-associated gas reserves were calculated from the
aggregated reserves discounting the associated gas production calculated until 2030.
Table 5.5 summarizes the EUR used for modelling in each region according to
the reserves/resources classification and the natural gas type.
96
Table 5.5. EUR in Mm3 of natural gas used in each region
Country Natural gas type
Reserves/ Resources Classification
Proven Probable Possible Other
Resources
Argentina
Associated natural gas 90 874 36 282 24 677 118 291
Conventional Non-associated natural gas 241 297 112 795 120 368 395 675
Unconventional Non-associated natural gas 5 978 256**
Bolivia Conventional Non-associated natural gas 295 944 99 120 117 528 662 103
Chile Conventional Non-associated natural gas 41 000 57 000 - 42 583
Unconventional Non-associated natural gas 69 458***
Integrated Brazil
Associated natural gas* 278 528 475 407 196 878 58 175
Conventional Non-associated natural gas 104 267 211 471 107 204 25 878
North Brazil Associated natural gas 23 660
Conventional Non-associated natural gas 30 389 58 848 *EUR for associated natural will be defined indirectly by the EUR of oil defined in Section 5.4.2 and
Section 5.4.3.
**Economically Recoverable.
*** Undiscovered fractile 95%.
Source: ANP (2015, 2016b); IEA (2015b); SECRETARIA DE ENERGIA (2015c);
USGS (2016); YPFB (2014)
As reckoned by MCGLADE (2013), estimating costs is a challenging task, not
only by the lack of public data but also for the different components considered when
companies report aggregated costs - for instance, if the reported costs include overhead
costs, government take, financial costs, etc.
As this thesis is within the framework of a conceptual study, many assumptions
have been taken for estimating the cost to have an accuracy of the order of magnitude
that can provide sensible results for the exercise.
In this sense, the aggregated approach based on the reserves and resources
categorization was used to estimate the escalated costs. CAPEX and OPEX ratios were
97
estimated based on historical investments reported by the different companies/agencies
and the historical natural gas production (ANP, 2015; EPE, 2014; PETROBRAS, 2016;
SECRETARIA DE ENERGIA, 2015b, 2015d; SGI, 2016; YPFB, 2015b, 2015c)12.
Costs were updated to 2012 using Producer Price Index (BLS, 2016). Based on
these cost ratios and the production profile described before, the annual CAPEX and
OPEX necessary to develop and produce the “Probable” Reserves for a typical field
were estimated. Then, this cost structure was the basis to estimate the CAPEX and
OPEX for each category of natural gas resource13
. Finally, to input the costs in the
TIMES model, the CAPEX related to production and exploration was discounted to
present value (year 0) as an “overnight investment cost” and OPEX was maintained at
nominal values as a “variable costs”(Figure 5.16).
12 Since there are no historical costs available of unconventional gas production in Magallanes
Basin in Chile, we used those estimated for Argentina but increased the well CAPEX according to the
depth. For Magallanes Basin we used an average depth of 12000 ft. for “Estratos con Favrella” shale
formation (EIA, 2013c), whereas Vaca Muerta shale formation has an average depth of 6500 ft(EIA,
2013d).
13 For more information, see the Cash-Flows developed in CHAVEZ-RODRIGUEZ et al. (2016a).
98
Figure 5.16. Discounting at present value the CAPEX procedure for the inputting
cost in the model.
Table 5.6 shows the estimated cost used for modelling the non-associated gas
fields in each country by classification of reserves/resources.
99
Table 5.6. Non-associated natural gas production costs adopted for the modelling.
ARGENTINA
Costs
(US$/boe)
Developed Undeveloped
Probable Possible
Other Other Unconventional
Proven Proven Resources F95 Resources
Mean Resources
CAPEX NPV a 2.63 2.89 4.53 6.77 13.21 11.75
Finding 0.00 0.00 0.40 1.12 2.24 2.44
Developing 2.63 2.89 4.13 5.65 10.96 9.32
OPEX
nominalc 4.80 4.80 5.70 6.55 7.63 10.07 7.63
BOLIVIA
Costs
(US$/boe)
Developed Undeveloped Probable Possible Other Resources
Proven Proven
CAPEX NPV a 2.20 2.65 3.66 4.87
Finding 0.00 0.00 0.21 0.60
Developing 2.20 2.65 3.45 4.27
OPEX
nominalc 1.50 1.50 1.78 2.04 2.38
INTEGRATED
BRAZIL
Costs
(US$/boe)
Developed Undeveloped Probable Possible Other Resources
Proven Proven
CAPEX NPV a 5.72 10.07 13.11 26.56
Finding 0.00 0.00 2.03 7.10
Developing 5.72 10.07 11.08 19.47
OPEX
nominalc 14.41 14.41 17.10 17.10 26.41
NORTH
BRAZIL
Costs
(US$/boe)
Developed Undeveloped Probable Possible Other Resources
Proven Proven
CAPEX NPV a 8.95 9.84 11.09 18.34
Finding 0.00 0.00 0.26 0.26
Developing 8.95 9.84 10.82 18.08
OPEX
nominalc 10.58 10.58 12.56 12.56 16.83
CHILE
Costs
(US$/boe)
Developed Undeveloped Probable Possible Other Resources
Unconventional
Proven Proven Resources
CAPEX NPV a 9.07 9.98 14.86 21.84 22.68
Finding 0.00 0.00 1.28 3.63 4.50
Developing 9.07 9.98 13.58 18.20 18.18
OPEX
nominalc 6.04 6.04 7.17 8.24 9.60 9.60
Source: Own estimations based on ANP (2015); EPE (2014); PETROBRAS
(2016); SECRETARIA DE ENERGIA (2015b, 2015d); SGI (2016); YPFB (2015b,
2015c)
100
5.3 Midstream Modelling
5.3.1 Natural Gas Processing Plants
Wet gas is transported to the treatment plant containing solid particles (fine
sands), liquid (mercury, oil, and natural gas heavy liquids), and harmful gases (CO2 and
H2S). The objective of a natural gas processing plant (NGPP) is to produce a methane-
rich gas and hydrocarbon liquids by removing the acid gases, nitrogen, water, and other
impurities(MAZYAN et al., 2016). Figure 5.17 shows the processes involved in the
natural gas treatment and fractioning in a NGPP. Wet gas is first dehydrated for the
removal of water to a sub-ppm level, which makes the gas acceptable for the
downstream cryogenic unit or for export requirements. Then, the dehydrated gas is
transferred to the acid gas removal unit (referred to as the gas sweetening unit) to
remove CO2 and H2S due to their corrosive properties. After this step, the sweetened
gas is transferred to mercury removal to prevent reaction of mercury with aluminium
heat exchangers. Nitrogen is removed in the Nitrogen Rejection Unit (NRU) to reduce
transportation volumes and increase heating value, where it is further dehydrated using
molecular sieve beds. Afterwards, methane is separated from natural gas liquids.
Cryogenic processing and absorption methods are some of the ways to separate methane
from NGLs and ethane. There are two main techniques for removing NGLs from the
natural gas stream: the absorption method and the cryogenic expander process. The
cryogenic method is better at extracting the lighter liquids, such as ethane. Finally
natural gas liquids are fractioned based on the different boiling points of the different
hydrocarbons in the NGL stream (EIA, 2006).
NGL, LPG and condensate as well as the pure components methane, ethane,
propane, and butane are often extracted and fractionated in tailor-made processing
plants according to the specific requirements of the regional market (BHRAN;
HASSANEAN; HELAL, 2016; LINDE, 2015). For our model, we considered a natural
gas processing plant with the following outputs: dry natural gas, ethane, LPG and plant
condensate/natural gasoline (C5+).
101
Figure 5.17. Natural Gas Processing Plant Schematic.
Source: Based on EIA (2006)
The CAPEX of NGPP will depend on the scale, the complexity and location of
the plant. Some plants such as Rio Grande and Gran Chaco in Bolivia receive the pre-
treated natural gas and are focused on the natural gas liquids separation. Other aspect is
the expansion of processing trains. For instance, in terms of costs per unit, the original
NGPP of Malvinas in Peru (440 Mcfd) has a cost per unit of capacity around 70% more
than cost per unit of capacity of the project after their second expansion (1680 Mcfd).
Figure 5.18 shows CAPEX estimations and capacities of different NGPP projects. As
we are not incorporating scale costs we adopted the 27.3 MUS$/Mcmd for CAPEX of
NGPP.
OPEX of NGPP is usually rated as a percentage of CAPEX costs. It ranges
between 1.5% and 4.0% of CAPEX(PETROBRAS, Brasilia, Brazil; SANTOS, R. M.,
2015). For modelling purposes, we adopted an OPEX of 3.0% of total CAPEX.
102
Figure 5.18.CAPEX estimations vs. Capacities of different Natural Gas
Processing Plants.
Natural gas is used in an NGPP as a fuel in equipment such as compressors and
reboilers. The energy demand in anNGPP ranges from 0.5% to 9.0% of produced gas
(SGI, 2015). Considering the risk of double accounting, since upstream natural gas
consumption reported is aggregated and might consider NGPP’s consumption, we
applied only 2% of self-consumption of dry gas in NGPP operations.
New NGPP reach a NGL recovery of over 99% and ethane recovery over 95%
(COSTAIN, [s.d.]; HUEBEL and MALSAM, 2012; LINDE, 2015). For modelling
purposes, considering the existing installed capacities, I adopted an ethane recovery
factor of 90% and heavier NGLs of 95%. The split of NGL into different products
depends on the NGL’s chemical composition. Data about the chemical composition of
the NGLs contained in the feed gas to NGPP is known only by operators and is rarely
publicly available. To overcome this lack of information, outputs per unit of NGL
fractionated based on Gran Chaco NGPP yields were adopted (FUNDACIÓN
JUBILEO, 2014)(Table 5.7). The installed capacities of natural gas processing plants
were obtained from (ANP, 2015; FUNDACIÓN JUBILEO, 2014; IPA, 2011; TGS,
2015).
y = 27.269x R² = 0.7601
0
200
400
600
800
1000
1200
1400
0 10 20 30 40 50
CAPEX (MUS$)
Capacity (MMcmd)
Malvinas ( Peru)
Gran Chaco ( Bolivia)
Caraguatatuba ( Brazil)
Rio Grande ( Bolivia)
103
Table 5.7. Yields adopted to model the fractionation process.
Input NGL 1 Tonne
Losses NGL 0.078 Tonne
Outputs
C2 0.496 Tonne
LPG 0.374 Tonne
C5+ 0.052 Tonne
.
5.3.2 Regasification and Liquefaction of Liquefied Natural Gas(LNG)
The phase transition occurs at -162°C reducing its volume to 1/600th the volume
of natural gas for storage and transport purposes. In the process of liquefying natural
gas, a conditioned stream is supplied to cryogenic LNG facility under elevated pressure,
it is then subjected to several cooling stages by indirect heat exchange with evaporating
refrigerants until it is completely liquefied. Most of the developed liquefaction
processes include what is commonly known as a precooling stage where natural gas
feed is cooled down to a temperature varying from −30 °C to −50 °C depending on the
precooling technology. After the precooling step comes to another cooling stage where
the precooled natural gas is condensed and subcooled; the pressurized LNG is then
flashed and stored at slightly above atmospheric pressure at approximately −162 °C
(UWITONZE et al., [s.d.]). According to HWANG and LEE (2014) the Propane pre-
cooled mixed refrigerant (Figure 5.19) is by far the most common technology used for
liquefaction of natural gas.
104
Figure 5.19. Diagram of a natural gas liquefaction process.
Source: HWANG and LEE (2014)
The conversion from liquid to gas can be done using onshore facilities or
regasification ships moored at specially designed docks. In the last case, the terminal is
based on a Floating Storage and Regasification Unit (FSRU) permanently moored at the
jetty and periodically supplied by an LNG carrier. This practice, called ship-to-ship
operation (STS), allows a continuous operation of the LNG import facility
(D’ALESSANDRO et al., 2016).
105
Figure 5.20. LNG Regasification Schematic Process.
Source: LEMMERS (2005)
LNG is transferred into a cargo tank by dedicated LNG feed pump. Then, LNG
is sent to a high-pressure booster pump, before entering anLNG vaporizer. Finally, LNG
is vaporized and transported to the end user. If required by the end user, odorant can be
mixed after the vaporization process (LEE et al., 2014).
In the last decade in the Southern Cone, seven LNG regasification plants have
been installed, only one onshore, the Quintero regasification plant in Chile. The other
six are FSRU. The main characteristics of these regasification plants are summarized in
the following table.
106
Table 5.8. Regasification Terminals in the Southern Cone.
First Year of
Operation
Name Region
Capacity (Million
m3 per day of
gas)
Storage
Capacity
(thousand
m3 of LNG)
Storage Capacity
(Million m3 of
gas)
2008
Bahia Blanca Gas
Port Argentina 17 151 94
2009 Quintero LNG
Central and South
Chile 10 334 207
2009 Pecém Brazil 7 129 80
2010 Mejillones LNG North Chile 5.5 175 109
2011 Puerto Escobar Argentina 17 151 94
2012 Guanabara LNG Brazil 20 173 107
2014 TRBA Brazil 14 137 85
2015
Quintero LNG
Expansion
Central and South
Chile 5 334 207
Source: CABANES (2015); GNL QUINTERO (2015); IEA (2015b); PETROBRAS
(2014b); RODRÍGUEZ (2011); YPF (2015)
For modelling purposes, investment costs adopted were: 150 US$/tonnes of
capacity per year for new regasification projects for imports of LNG; 1400 US$/tonnes
of capacity per year and 1800 US$/tonnes of capacity per year for onshore and floating
liquefaction plants for exports of LNG, respectively (IGU, 2015). Operating costs were
considered as 4% of the capital cost per year for both technologies. Based on IEA-
ETSAP (2011) losses of 2.5% and 11% were considered for regasification and
liquefaction plants, respectively.
Defining LNG prices is a critical issue for the model, since it can determine if it
is less costly to import LNG or tap the domestic resources to supply the domestic
demands. Figure 5.21shows the historical average CIF prices of LNG imports in Brazil,
Chile and Argentina from 2010 to the first quarter of 2016. In most of the recent years,
Chile benefited from lower average import prices compared to Argentina and Brazil.
This is because Chile has long-term “take-or-pay” contracts with its suppliers.
107
For the sake of simplicity, the same LNG import prices of the first quarter of
2016 will be maintained for each country until 2030. In the case of exports for
liquefaction projects, a FOB price of 5 US$/MMBtu is applied over the time horizon.
In chapter 7, a sensitivity analysis changing the LNG prices is performed to tackle the
uncertainty of LNG prices evolution in the future.
Figure 5.21. Average prices of LNG imports in the Southern Cone.
Source: CNE (2016c); MINEM (2016); MME (2016b)
5.3.3 Gas Pipelines
The transportation segment of the natural gas industry is responsible for
delivering natural gas from domestic producers or imports to the market areas via
pipelines. The transportation system is composed of pipelines, compressor stations,
citygate stations, and storage facilities. The main differences among the pipeline
systems are their physical properties (e.g., diameter, stiffness and material) and the
specifications of their maximum and minimum upstream and downstream pressures.
Gas pipelines do not only serve as transportation links between producer and
consumer, but they also represent potential storage units for safety stocks. That is, due
to the compressible nature of dry gas, large reserves can be stored on a short-term basis
inside the pipeline through a process called line packing. This is accomplished by
0
2
4
6
8
10
12
14
16
18
2010 2011 2012 2013 2014 2015 2016-Q1
US$/MMBtu
Brazil
Chile
Argentina
108
injecting more gas into the pipelines during off-peak times by increasing the gas
pressure, and by withdrawing larger amounts of gas during periods of high demand
when flow capacities elsewhere in the system break down (RÍOS-MERCADO and
BORRAZ-SÁNCHEZ, 2015).
In the work of CHAVEZ-RODRIGUEZ et al. (2016b), an hourly approach was
used both for the demand and the supply. As in that work, the line packing was not
considered the results did not match the real operation of the system. For instance,
without considering line pack, in CHAVEZ-RODRIGUEZ et al. (2016b), the
production wells and natural gas processing plants had a high fluctuation on an hourly
basis as the demand, instead of working under a regular regime as it is observed in the
real world. This is one of the reasons why the midstream in this thesis was modelled on
a monthly basis: to avoid the modelling of the line pack.
In TIMES-ConoSur, only the existing international pipelines infrastructure was
modelled. Neither domestic transmission nor distribution pipelines were considered as
this is out of the scope of this regional study. These international natural gas pipelines
were modelled as a total transportation capacity across regions and a unidirectional flow
process (in chapter 7 was performed a sensitivity analysis including bidirectional flow
between Argentina and Chile). Table 5.9 shows the international pipelines modelled.
109
Table 5.9. International natural gas pipelines in the Southern Cone.
IMPORTERS
International Pipelines
Argentina Bolivia Brazil North Chile Central and South Chile
EX
PO
RT
ER
S
Argentina
Aldea Brasileña-Uruguaiana: 12 Mm3/d. Flows interrupted in 2008.
Norandino: 5 Mm3/d. Flows
interrupted in 2008. Atacama: 9
Mm3/d. Flows interrupted in 2008.
GasAndes: 9.5 Mm3/d. Gasoducto del Pacífico: 3.5 Mm3/d . Methanex I: 2 Mm3/d. Methanex II: 2.8 Mm3/d. Methanex III: 2 Mm3/d. For all these pipelines the natural gas flow was interrupted in 2008.
Bolivia
Yabog: 6.5Mm3/dia. Currently working as contingency pipeline. Madrejones-Campo Duran: 1.2 Mm3/d. Abandoned in 2012. Juana Azurduy de Padilla(GIJA) : 24 Mm3/d. In operation since 2011, it is planned to be expanded until 33 Mm3/d in 2019.
Gasbol: 30 Mm3/d. At full capacity. Cuiaba: 2.6 Mm3/d. At full capacity.
Brazil
North Chile
Central and South Chile
Recently, Argentina signed an agreement with Chile to import natural gas in
the winter season using the Gas Andes and Norandino pipelines infrastructure and the
regasification capacity of Quinteros and Mejillones plants. The agreement is for 3
Mm3/day through the Gas Andes pipeline, including the possibility to increase it by 1
Mm3/day, and by 1 Mm3/day through the Norandino pipeline. These transfers already
started in the winter of 2016. For modelling purposes from 2017 on, we will not limit
the transfers by the commitments defined in Gas Andes and Norandino pipelines
agreements. The transfers will be limited by the nominal capacity of these pipelines, in
order to analyze the effects of the trade between Argentina and Chile under the least
cost operation.
110
5.4 Scenarios
The primary goal of energy planning tools are the insights that quantitative
modelling provides rather than the values of precise-looking projections they can
produce in a given scenario (HUNTINGTON et al., 1982). As the purpose of this thesis
is to provide insights on how the natural gas markets in the future can evolve and to
assess what role the factors addressed in Section 2 play in this future, the analysis is
underpinned on the construction of scenarios.
Scenarios can assume different roles in foresight exercises and this diversity has
its impacts on the role and usefulness of the proposed context scenario approach
(WEIMER-JEHLE et al., 2016). According to BÖRJESON et al. (2006) scenarios can
be‘predictive’ (e.g. what-if or what will happen), ‘explorative’ (what may happen or
what can happen) or‘normative’ (what should happen or how can a specific target be
reached).
Explorative scenarios aim to explore situations or developments that are regarded
as possible to happen. Explorative scenarios resemble “what-if” scenarios and are
elaborated with a long time-horizon to explicitly allow for structural, and hence more
profound, changes. Explorative scenarios are mainly useful for analysing strategic
issues (BÖRJESON et al., 2006).
Despite the nature of TIMES as a normative scenarios model (PFENNINGER et
al., 2014), inputting scenarios constraints into the model allows building either
predictive or explorative scenarios. Therefore, it is relevant to define the constraint(s),
which have major impacts on the evolution of the energy system studied, in the case of
this thesis the natural gas markets system in the Southern Cone.
Two mains scenarios were constructed that are structurally different in the
evolution of natural gas markets in the Southern Cone: The “Constrained Investment
Scenario” and the“Unconstrained Investment Scenario”, the main constraints inputted in
the model to build these scenarios are indicated in Table 5.10. These constraints were
applied only for Argentina and Brazil.
111
Table 5.10. Summary of the main scenarios simulated in this thesis
Countries / Scenarios Constrained Scenario Unconstrained Scenario
Argentina Constraints on the total
CAPEX investments in
non-associated gas
production from 2016
onwards.
Without constraints from
2016 onwards.
Bolivia Without constraints Without constraints
Brazil Conservative offshore oil
production curve
Optimistic offshore oil
production curve
Chile Without constraints Without constraints
TIMES-ConoSur allows inputting exogenously limitations on the total
investments per year for a specific or a group of technologies. This approach was
applied in Argentina for all non-associated gas fields (also unconventional gas fields)
where CAPEX costs were specified. As the levels of investment per year in Argentina
are difficult to predict, the Constrained Investment Scenario includes arbitrary
conservative assumptions that lead to low natural gas production to test how the natural
gas markets in the Southern Cone would evolve under this situation.
The methodology used for Brazilian offshore oil production where associated gas
is produced was top-down methodology, a Multi-Hubbert curve. The use of this
methodology avoids including costs of oil fields. Therefore, in this top-down approach,
each scenario included different EUR for the Multi-Hubbert curves to reflect indirectly
the restriction on the capacity of investments.
For both Constrained and Unconstrained Scenarios, the expansion of
regasification plants for LNG is only allowed from 2018 onwards, to take into account
the engineering and construction time.
112
In the next section, it is explained why the capacity of investments was pivotal to
define the main scenarios in this study. In section 5.4.2 and 5.4.3 further modelling
details are provided about the Constrained and Unconstrained Scenarios respectively.
These main scenarios are the basis for further sensitivity scenarios elaborated in chapter
7.
5.4.1 Investment capacity in upstream as the main constraint for the future natural gas
market in the Southern Cone
Previous modelling experiences using TIMES-ConoSur (CHAVEZ-
RODRIGUEZ et al., 2016b, 2016c) found that the capacity to develop capital
investments in upstream is pivotal for the future supply of natural gas in the Southern
Cone and will shape the future of LNG and regional gas trade. This was an outcome of
a purely modelling exercise. In this section, it is explained what is referred to
investment capacity and its theoretical framework in the Southern Cone to underpin the
modelling assumptions for the Constraint and Unconstraint Scenarios.
The limits to invest in upstream might be explained by financial constraints, lack
of industrial capacity and regulatory constraints.Capital expenditures in upstream of oil
and gas companies are being reduced worldwide due to currently low oil prices (OGJ,
2016) and the Southern Cone is not an isolated case. The output of the TIMES model
includes investment decisions (e.g. an oil field with a break-even price higher than the
oil import price will not be developed), which can be made freely or under constraints.
The later can be inputted exogenously to reflect the challenges to attaining funding for
investments in upstream.
Funding mechanism for upstream oil and gas investments depend on the type of
company involved and the type of project. For instance, for Majors14
the income
generated by their portfolio of upstream assets has traditionally been the main source of
funding for CAPEX. This can be supplemented,if necessary, by corporate borrowing
14The “Majors” are a group of multinational oil companies given the moniker due to their size, age
or market position, such as BP, Shell, ExxonMobil, etc..
113
from banks or from the capital markets(IEA, 2014a). On the other hands, some NOCs
such as Petrobras, because of the speed at which it plans to increase its production in the
coming years, relies more on debt financing for its capital spending programme (IEA,
2014a).
Petrobras’ financial difficulties have been stressed due to the fuel price policy
adopted by the Brazilian government and to problems related to corruption scandals,
undermining the rate of investments in the Brazilian Oil Industry (RODRIGUES, N.;
COLOMER, 2016). This lead to a reduction inPetrobras investments plan from US$
236.7 billion (PNG 2013-2017) to US$ 130.3 billion (PNG 2015-2019) (PETROBRAS,
2013, 2015b). As a matter of fact, Petrobras is implementing a disinvestment program
(estimated in US$ 15.1 billion for the period 2015-2016) contemplating the sale of
minority, majority or entire positions in some subsidiaries, affiliates, and assets in order
to meet the funding needs to leverage deep-water projects in Brazil (PETROBRAS,
2016).
In the case of Argentina, oil and gas companies are largely dependent on
economic conditions and the country’s institutional framework. CAPEX and OPEX
costs are subject to the level of inflation. In addition, Argentina imposed currency
exchange controls and transfer restrictions, which limited the ability of companies to
retain foreign currency or make payments abroad. In spite of the relief of these
restrictions by the current administration, there is no assurance that future regulatory
changes related to exchange and capital controls will be imposed again. This does not
only constraint the ability to finance and pay the planned CAPEX in foreign currency,
but also affects the interest rates because of the risks (HARDEN, 2014).Companies such
as YPF are financing their CAPEX expenses relying on borrowing funds in dollars and
sensitive to changes in interest rates because of Argentina’s country risk (YPF, 2016).
Another factor affecting financial capacity for oil and gas companies in Argentina
is that it is mandatory to prioritize the domestic market where prices are controlled. As
stated by YPF: “...We budget capital expenditures by taking into account, among other
things, market prices for our hydrocarbon products… Our ability to execute and carry
out our business plan depends upon our ability to obtain financing at a reasonable cost
114
and on reasonable terms…” (YPF, 2016). This will be crucial,especially for tapping
unconventional resources (DI SBROIAVACCA, 2013).
Additionally, some of the customers of oil and gas companies in Argentina are
government entities, which struggle to pay for their consumption. As an example, at the
end of 2015, the accounts “receivable balance” corresponding to the “Natural Gas
Additional Injection Stimulus Program” for YPF indicated nine months of accrued,
unpaid payments, representing approximately US$ 1.1 billion15
(YPF, 2016).
In Bolivia, as described by CHAVEZ-RODRIGUEZ et al. (2016a), despite to the
good export prices for natural gas and liquids, the oil and gas companies have to supply
first the domestic market where prices are subsidized. This, together with a high
government take (in average 50% without taking into account corporate taxes) is
diminishing the net profits of the oil and gas companies and consequently their financial
capacity. As stated in Section 3.2.2, taxes and subsidies were not incorporated in this
long-term study despite its relevance. For the sake of simplicity, taking into account the
low-cost of Bolivian natural gas and its good export price, it is assumed that the
regulatory framework will be improved16
and oil and gas operators in Bolivia, specially
YPFB, will not have the same financial limitations than operators in Argentina and
Brazil.
Finally, in Chile the investments required in upstream for tapping resources in the
Magallanes Basin are much smaller,when compared to the other countries assessed, and
credit risk for ENAP is low (ENAP, 2016). As suggested by ROJAS (2012) the
production of petroleum resources in Chile will rely mainly on their economic
feasibility. Therefore, no financial constraints are assumed for this country in the
modelling exercise.
15 9.9 billion Argentinian Pesos (Ps.) using an average exchange for 2015 of 9.3 Ps./US$(BCRA,
2016).
16Actually, at the end of 2015, the Government of Bolivia has issued Law No. 767 that introduces
price incentives for oil production and for condensates produced in new gas fields discovered.
115
5.4.2 Constrained Investment Scenario
Investments in CAPEX in upstream worldwide were estimated to fell around 40-
50% in 2016 when compared to 2014, due mainly to low oil prices (SGI, 2016). To
capture in the model this fact for Argentina it was projected a decline of 50% in the
CAPEX investments for 2016 when compared to 2014 (SECRETARIA DE ENERGIA,
2015d). From 2016 onwards the Constrained Investment Scenario considers a CAPEX
investments decline of 3% per year in non-associated conventional gas projects and a
growth of +15% per year in non-associated unconventional gas projects17
(Figure 6.1).
Figure 5.22. Investments in upstream CAPEX per year in Argentina considered
for the Constrained Investments Scenario.
For Brazil, in the Constrained Investment Scenario, P95 reserves (Table 5.5) were
used as the EUR for Post-salt oil production and 30 billion oil barrels as the EUR for the
Pre-Salt oil production. The results of oil production projections using a Multi-Hubbert
methodology with the EUR specified are showed in Figure 5.23.Furthermore, the
natural gas associated with this oil production was estimated using the natural gas to oil
ratios specified in Figure 5.12.Parameters of the Brazilian Multi-Hubbert of oil
production can be found in Table 10.8 in Annex B
17A higher growth in unconventional resources investments than in conventional resources is based
on the growth pace of the last yeras(SECRETARIA DE ENERGIA, 2015d).
0
500
1000
1500
2000
2500
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
MMUS$ Argentina
Conventional Gas Total CAPEX Unconventional Gas Total CAPEX
-50%
-3%
+15%
116
Figure 5.23. Brazil’s oil production curve for the Constraint Investment Scenario.
Figure 5.23 shows that Post-Salt production maintains a declining pattern and the
growth in production of Pre-Salt oil fields are not enough to increase the oil production
in the country. Only from 2020 the oil production rises again reaching 2.5 MMbpd in
2030.
5.4.3 Unconstrained Investment Scenario
This scenario also incorporated the decrease in CAPEX investment in 2016
.However, from 2016 on there is no restriction in CAPEX for Argentina. This
unconstrained capacity of investment aims to test what may happen if funding and
industrial capacity were adequate to tap the conventional and unconventional gas
resources in Argentina.
0
1 000
2 000
3 000
4 000
5 000
1954
1959
1964
1969
1974
1979
1984
1989
1994
1999
2004
2009
2014
2019
2024
2029
2034
2039
2044
2049
2054
2059
2064
2069
2074
2079
2084
kbpd
Post-Sal Production Pre-SaltOnshore Offshore <400mOffshore >400m H1 Offshore >400m H2Multi-Hubbert Post-Salt (P50) Pre-Salt Hubbert 100MH Post-Salt + MH Pre-Salt
2030
117
For Brazil, in the Unconstrained Investment Scenario, P50 reserves (Table 5.5)
were used as the EUR for Post-salt oil production and 100 billion oil barrels as the EUR
for the Pre-Salt oil production. Figure 5.24 shows the results using this EUR on the
Multi-Hubbert curves for Brazilian oil production.
Figure 5.24. Brazil’s oil production curve for the Unconstrained Investment
Scenario.
Figure 5.24 shows that in 2030 in the Unconstrained Investment Scenario the
Brazilian oil production reaches nearly 6 Mbpd. Comparatively, this is twice the oil
production in the Constrained Investment Scenario. Differently from the Constrained
Investment Scenario, in this Scenario production in Post-Salt fields increase along with
higher production levels at Pre-Salt fields.
It is important to highlight that these are two arbitrary scenarios with a set of
assumptions adopted to test the TIMES-ConoSur model aiming to provide insights
about different futures of the Southern Cone upstream and its impact on the natural gas
markets. The outcomes of the model and findings should be encompassed in this limited
framework.
0
1 000
2 000
3 000
4 000
5 000
19
54
19
59
19
64
196
9
19
74
19
79
19
84
19
89
199
4
19
99
20
04
20
09
20
14
201
9
20
24
20
29
20
34
20
39
204
4
20
49
20
54
20
59
20
64
206
9
20
74
20
79
20
84
kbpd
Post-Sal Production Pre-SaltOnshore Offshore <400mOffshore >400m H1 Offshore >400m H2Multi-Hubbert Post-Salt (P50) Pre-Salt Hubbert 100
2030
118
6 Results
This section presents the results of TIMES-ConoSur according to the two
scenarios defined in Section 5.4. These findings focus on:
- Natural gas production: where the gross production of natural gas at a
yearly resolution before the self-consumption and losses in upstream is
shown.
- Natural gas supply: where the supply side of the market is shown by the
domestic production of natural gas, imports by pipelines and
regasification plants at a monthly level.
- Power generation: where the electricity produced by different
technologies at an hourly resolution is shown.
- Natural gas demand: where the natural gas consumed for power
generation and end-uses at a monthly level is shown.
6.1 Natural gas domestic production results
On the production side, the model outputs provide useful insights about when and
how much natural gas resources would be exploited and depleted according to each
category.Gross natural gas domestic production under in the Constrained Investments
Scenario and the Unconstrained Scenarios are presented in Figure 6.1 and Figure 6.2
respectively.
119
Figure 6.1. Gross natural gas domestic production under the Constrained
Investments Scenario
120
Figure 6.2. Gross natural gas domestic production under the Unconstrained
Investment Scenario
In a Constrained Investment Scenario Argentina continues its decline of natural
gas production decreasing to 84% of 2012’s gross production in 2030. Conventional gas
remains as the major supply source. In turn, in the Unconstrained Investment Scenario,
in the first-year with no investments limitation (2017), the production boosts relying on
conventional gas possible reserves and other resources until 2017 when occurs the peak
production of conventional gas. From 2018 onwards the natural gas production expands
relying on unconventional resources, which would account for 63% of total gas
production in the country in 2030. In this scenario the natural gas production in 2030 is
27% greater than in 2012.
121
For Bolivia, both scenarios are almost similar despite the changes in the supply
dynamics of its natural gas importers. In the Unconstrained Scenario the accumulated
production up to 2030 is 2% lower than the Constrained scenario. This can be explained
by the low-cost of its natural gas. In both scenarios, 3P reserves are already committed
in 2021, from 2022 onwards the expansion of production relies on other resources.
In Brazil, both Pre-Salt and Post-Salt associated natural gas plays a major role in
the supply of this resource for the country. In the Constrained Scenario, provided the
decline rates of Post-Salt fields production continues and despite the increase of Pre-
Salt fields production, the total natural gas production follows a declining pattern until
2022 when the growing effect of Pre-Salt Fields overcomes the declining effect of both
Post-Salt Fields and non-associated developed gas fields. In 2030 in the Constrained
Investment Scenario the production level of associated gas is 40% larger than in 2012.
Interestingly, in this scenario, undeveloped non-associated gas reserves are extracted in
the second half of the next decade. This does not occur in the Unconstrained Scenario
due to the flood of associated gas. In this case, the associated gas production in 2030 is
178% higher than the 2012 production level.
In the North of Brazil, natural gas production remains low this decade in both
scenarios. However, in the 2020’s, non-associated gas is triggered in both scenarios at
different levels. As this chapter will detail further, this is explained by the consumption
in thermal power plants. North Brazil gross-production in the Constrained investment
Scenario nearly increases by a half the production of the Unconstrained investment
Scenario. This is explained by the less associated gas production available for power
generation in the Constrained scenario in the Integrated Brazil region and the cost-
differential of non-associated gas production between North and Integrated Brazil.
In Chile, the production does not vary significantly between the two scenarios.
The natural gas production is triggered in the first-year of simulation (2017) and in the
“Constrained” scenario a higher level of production is slightly anticipated. It is relevant
to say that the model did not choose to produce the high-cost unconventional resources
in Chile, despite the lack of constrain to the capacity of investments. Instead, to supply
domestic demand it opted to deploy LNG regasification plants. These are the kind of
122
economic insights given by a model aiming the minimization of the system costs, which
can support national energy supply strategies.
6.2 Natural gas supply results
Figure 6.3 and Figure 6.4 shows the supply sources of natural gas for each region,
including domestic production, LNG and pipelines imports. In general, results
corroborate the strong dependence and interaction between the regions evaluated in this
thesis. Bolivia remains as the main supplier in the region. Should the Unconstrained
Scenario occur, Argentina retakes natural gas exports to Chile, with higher volumes sent
to the Central-South Chile instead of the North Chile region. No exports from Argentina
to Brazil are expected to happen if supply commitments are not considered. Despite its
natural gas exporter position under an unconstrained scenario, Argentina should keep its
LNG imports in both scenarios during the winter season. The domestic production level
of the different scenarios determines the degree of dependence of LNG for Argentina.
New LNG regasification plants in Argentina are required in 2018, the first year when
they are allowed by the model.
Moreover, the modelling of the power sector in this study neglected the effects of
El Niño Southern Oscillation, which affects severely the hydropower output, therefore
resultin in a dramatic increase of LNG imports during dry-years, as observed
historically. In addition, other factors such as unavailability of power plants,
congestions of transmission lines, commercial issues (i.e. contracts obligations),
contingencies, flexibility in the operation, etc. were not taken into account in this
modelling exercise. These factors would likely upwards the requirements of LNG in the
region, even in those years when the TIMES-ConoSur model provides zero LNG
imports.
123
Figure 6.3. Natural gas supply projections in the Southern Cone under the
Constrained Investment Scenarios
124
Figure 6.4. Natural gas supply projections in the Southern Cone under the
Unconstrained Investment Scenarios
Chile also requires expanding its LNG regasification capacity in both scenarios.
However, in the Unconstrained Scenario, the volumes of LNG are much lower due to
the Argentinian natural gas imports in the summer. Interestingly, the results of Chile
suggest that Argentinian exports to Chile are higher in the first ten years when low-cost
resources produced in Argentina have lower break-even prices than imports of LNG in
Chile. In North Chile current regasification capacity matches the regional needs in this
decade. This brings opportunities to use the spare regasification capacity in Chile for
exports to Argentina.
125
In the Constrained Scenario, LNG in Brazil is not required until 2022 if there are
regular inflows for hydropower plants that provide the capacity factor assumed for this
study and the new power plants start operation as planned (Figure 5.2). Imports of LNG
in this scenario accentuates in Brazil during the dry-seasons. These figures indicate the
strong complementarity of LNG imports and hydropower seasonality. None of the
scenarios requires the expansion of regasification capacity in Brazil until 2030. Even, in
the Unconstrained Scenario, no further LNG imports are required after 2015.
Alternatively to LNG, the model suggests that swing producer petroleum fields
are required to increase production in dry-seasons. This is a monthly allocation of the
model for the associated natural gas inputted on a yearly resolution. Of course, oil fields
do not follow a swing producer pattern just to allocate associated-natural gas. Therefore
without natural gas storage, this supply pattern for dry-season does not occur. This
remarks the necessity of developing natural gas storage in the model to incorporate
costs and to be able to input associated gas with a regular production pattern under a
monthly resolution This is discussed with further detail in section 7.
In the North Brazil Region, specifically, the Cuiaba Thermal power plant reduces
slightly its consumption of Bolivian natural gas in the Unconstrained Scenario. This is
explained by the fact that other natural gas power plants operate in the Integrated Brazil
region using low-cost associated gas produced.
6.3 Power Generation Results
Power generation results in intervals of 5 years from 2015 until 2030 are shown in
Figure 6.5 and Figure 6.6 for the Constrained and Unconstrained Investment Scenarios
respectively. In general, with the exception of North Brazil, there are no relevant
differences between scenarios in terms of the operation of the power sector and the
participation of natural gas, this means that, for the prices of LNG assumed in the model
either with supply of LNG or domestic production, the optimal operation of the power
generation is quite similar in both scenarios. The model results reveal the role of natural
gas in power generation as a flexible and reliable source, particularly in peak hours.
126
However, in the case of Argentina and Bolivia, natural gas power plants supply a
significant part of the base load. In addition, power generation based on natural gas is
expected to increase. Nuclear and Wind increases significantly in Argentina as well18
.
During winter, the model operates natural gas power plants at a lower level than during
summer, reflecting the real seasonal pattern observed inFigure 2.3. Finally, in
Argentina, electricity imports from Brazil can be observed as a marginal supply during
the wet-season of Brazil, January to June, using the existing interconnections.
In the case of Brazil, seasonality and the increase of hydropower in North are
remarkable. Dry-season is defined from June to December for hydropower in this
region. As can be observed in the figures for Integrated Brazil, the electricity transfers
from North Brazil are higher in the wet season. Therefore, combined cycle plants
operate in the dry season in Integrated Brazil with higher capacity factors.
Coal increases its participation in Chile. Solar plays a major role as well. The
hourly resolution in the power sector of the model allows showing the operation of solar
power plants during the day and how they share with natural gas power plants part of
the peak hours. In Central-South Chile natural gas power plants have a larger operation
in the dry-seasons (from February to August). Interestingly, the capacity of coal plants
matches the North Chile demand. However, after the interconnections between SIC and
SIGN in 2018, the model runs combined cycle plants in North Chile to export electricity
to the Central-South region.
It is relevant to take into account that the methodology applied for power
generation involved the adoption capacities by technologies and electricity demand
projections of national power expansion plans. The objective of these plans is to
guarantee the energy supply, in this sense, most of them overestimate the demand of
electricity and the power capacity required to attend it, therefore, the natural gas
consumption and LNG estimated in this study is likely to be underestimated.
18 The power generation expansion plan adopted was the same for both scenarios, nevertheless it is
unlikely that in a scenario with a lack of capacity for investment in the upstream of oil and gas in
Argentina nuclear power plants projects can be developed.
127
Figure 6.5. Power generation profiles in the Southern Cone under the Constrained
Investment Scenario
128
Figure 6.6. Power generation profiles in the Southern Cone under the
Unconstrained Investment Scenario
129
6.4 Natural Gas Demand
Projections for the end-use and power generation natural gas demands are shown
in both Figure 6.7 and Figure 6.8.
The end-use demand by regions is assumed to be inelastic and is the same for
both-scenarios developed in this thesis. Industrial growth in all Southern countries
causes the increase of the natural gas consumption from 1040 PJ in 2012 to 1917 PJ in
2030 (a growth of 3.5 p.y.%). As indicated by CHAVEZ-RODRIGUEZ et al. (2016a) in
the case of Bolivia, this is part of a government strategy of “industrialization” of natural
gas through petrochemical plants, aiming to add value to this resource. In Chile, the
natural gas consumption also boosts because of the repressed consumption of natural
gas in the industrial and transport sectors.
The transportation sector also contributes significantly to the growth of
consumption, from 194 PJ in 2012 to 334 PJ in 2030 (3.1% p.y.). Summing all the
buildings sector, the projected consumption in the Southern cone grows from 506 PJ in
2012 to 786 PJ in 2030 (2.5% p.y.).
In Argentina and Chile, the natural gas demand for the residential and commercial
sectors incorporates the seasonality pattern of peak consumption in winter associated
with heating loads. In the Unconstrained Scenario, in the summer season, Argentina is
able to export natural gas to Chile, and this external market allows Argentina to
continue producing natural gas close to nominal capacity level (see Figure 6.8).
Figures for Bolivia highlight the relevance of exporting markets compared to the
country’s demand. Differences between scenarios are projected to happen in the second
half of the next decade, according to Figure 6.3 and Figure 6.4. These changes are
explained by different consumptions in North Brazil and Argentina, however, they are
not significant.
130
Figure 6.7 also evidences a higher consumption in North Brazil in a Constrained
Scenario (1203 PJ accumulated up to 2030) compared to the consumption in the
Unconstrained scenario (1073 PJ accumulated) due to a lower availability of natural gas
for power generation in Integrated Brazil. This shows that if less associated offshore
natural gas is produced, then, natural gas resources from Amazonia are necessary to
supply the power generation demand.
Finally, relying on the Argentinian gas imports during the summer season in the
Unconstrained Scenario, the Central-South Chile will have a higher consumption of
natural gas for power generation. The intermittent consumption pattern of natural gas
for power generation in North Chile is explained by the high share of coal in its
electricity generation mix. Actually, natural gas is consumed when “spot” monthly price
(shadow price) makes the operation of combined cycle plants to export electricity to the
Central-South region feasible (Figure 6.5 and Figure 6.6)19
.
The total natural gas demand summed of end-users and power generation in the
Constrained Scenario is projected to increase from 2794 PJ in 2012 to 4409 PJ in 2030,
whereas in the Unconstrained Scenario it reaches a consumption of 4448 PJ in 2030.
Differences between accumulated consumption in this time horizon sum 825 PJ; this
represents 2% of the accumulated consumption in the Unconstrained Scenarios (66 343
PJ). This small difference means that for the planning of the natural gas supply
infrastructure in a perfect foresight condition, LNG could substitute domestic
consumption without affecting significantly natural gas demand for power generation
under the prices of LNG assumed.
19According to the outputs of the model, natural gas power plants in North-Chile are not used for peaks in the
North region since this region has a “flat” load shape (see Figure 5.6).
131
Figure 6.7. Natural gas demand projection in the Southern Cone under the
Constrained Investment Scenarios
132
Figure 6.8. Natural gas demand projection in the Southern Cone under the
Unconstrained Investment Scenarios
For additional information, Table 6.1 provides the Natural gas supply and demand
balance for the Southern Cone. In both scenarios, Non-Associated gas represents the
major source of natural gas, 74% of the total domestic gross production in the
Constrained scenario compared to 67% in the Unconstrained scenario. In the
unconstrained scenario despite the increasing unconventional non-associated gas in
Argentina, the impact of the Brazilian offshore associated gas is greater at the end of the
period assessed. As can be expected that the volume of gas lost and consumed in
upstream increases with a higher gross production in the unconstrained investment
scenario, however is remarkable that in this scenario, as there is a flood of associated
gas in Brazil, the model chooses to re-inject the associated gas because there is no
133
market for it (a specific analysis of this is made in section 7.4). The accumulated LNG
imports in the Constrained scenario (14214 PJ) is more than double that of the
Unconstrained scenario (6481 PJ) with the LNG prices assumed (Figure 5.21). A
sensitivity analysis is made for the LNG imports in Section 7.2. Accumulated natural
gas trade within the Southern Cone borders increases 15% in the unconstrained
scenario, explained mainly by the retaking of Argentinian natural gas exports to Chile.
Imports of natural gas from Bolivia were maintained based on the assumption that
contracts will be renewed with the same take-or-pay features, an unconstrained analysis
is made in section 7.4 regarding international transfers. Finally, total accumulated
demand was not affected significantly between these two scenarios, unconstrained
scenario presents a total accumulated gas demand 1.2 % higher than the constrained
scenario (65518 PJ). These insignificant differences, as explained before, are due to the
current gap prices assumed between the domestic production and LNG imports, which
does not affect significantly the choice of natural gas technologies in power operation.
An additional insight is provided in section 7.1 regarding renewables and hydropower.
134
Table 6.1. Natural Gas Balance modeled in the Southern Cone for different
Scenarios
Scenario Category 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
Constrained Investment
Associated Gas (+) 898 1062 1175 1114 1033 992 962 942 928 920
Non-Associated Gas (+) 2355 2473 2407 2345 2337 2381 2536 2615 2628 2624
Upstream Losses and Consumption (-) 409 440 439 414 392 383 395 399 393 385
Additional Reinjection (-) 302 475 443 320 156 17 0 0 0 0
Net Wet Gas (=) 2542 2620 2699 2725 2822 2973 3102 3159 3164 3158
NGPP Losses (-) 48 50 51 52 54 57 59 60 60 60
NGL (-) 140 142 137 134 133 140 150 153 150 146
Dry Gas (=) 2354 2428 2511 2539 2636 2777 2893 2946 2953 2952
LNG Imports (+) 440 671 299 421 536 598 868 1052 1208 1458
Imports by Pipelines (+) 525 636 719 730 749 778 783 784 780 783
Exports by Pipelines (-) 525 636 719 730 749 778 783 784 780 783
Total Demand (=) 2794 3099 2810 2960 3172 3375 3761 3998 4161 4410
Power Generation (-) 1055 1243 871 901 969 992 1183 1276 1284 1372
Demand for End-Uses (-) 1739 1857 1939 2059 2204 2383 2578 2722 2877 3038
Total Discounted System Costs (MUS$) 34263
Unconstrained Investment
Associated Gas (+) 898 1062 1253 1256 1268 1307 1373 1446 1518 1588
Non-Associated Gas (+) 2355 2473 2397 2793 2850 2878 3073 3236 3352 3223
Upstream Losses and Consumption (-) 409 440 452 504 511 516 548 580 606 594
Additional Reinjection (-) 302 475 511 442 333 201 95 94 144 102
Net Wet Gas (=) 2542 2620 2687 3102 3274 3468 3804 4008 4120 4115
NGPP Losses (-) 48 50 51 59 62 66 72 76 78 78
NGL (-) 140 142 136 162 167 167 189 199 203 193
Dry Gas (=) 2354 2428 2500 2881 3045 3235 3543 3734 3838 3843
LNG Imports (+) 440 671 305 154 181 216 280 321 371 605
Imports by Pipelines (+) 525 636 718 869 883 942 940 957 929 841
Exports by Pipelines (-) 525 636 718 869 883 942 940 957 929 841
Total Demand (=) 2794 3099 2805 3035 3226 3451 3822 4054 4209 4448
Power Generation (-) 1055 1243 866 976 1023 1068 1245 1332 1332 1410
Demand for End-Uses (-) 1739 1857 1939 2059 2204 2383 2578 2722 2877 3038
Total Discounted System Costs (MUS$) 34653
135
7 Discussions
The main outcomes of the model were depicted in Chapter 6, providing findings
on how the natural gas supply and demand markets in the Southern Cone can evolve
under different investment scenarios.
In this chapter each of the specific research questions presented in Chapter 1 is
discussed based on the results provided by the TIMES-Conosur model. Moreover,
additional simulations were performed to have sensitivity outputs for each topic.
7.1 What is the economic potential ofArgentina’s unconventional?
Considering the 801.5 Tcf wet shale gas technically recoverable estimates of the
EIA (2013d) for Argentina, it is paramount to have estimates of how much Argentina is
able to economically produce when constrained by their resources costs and the market
for these resources. Figure 7.1 shows the production of unconventional gas projected for
Argentina in the Constrained and Unconstrained Investment Scenarios. In the
Constrained Investment Scenario, the unconventional gas production can reach 18
Mm3/d in 2030. This is only 61% more than the records of 2014 for unconventional
production. Hence, this represents a conservative perspective on investment in
unconventional gas in Argentina, which intends to incorporate the regulatory challenge
that upstream operators in Argentina faces.
On the other side, in the Unconstrained Scenario, the unconventional production
reaches 85 Mm3/d in 2030, similarly to the total of non-associated gas estimated in
2015 for this country (84.6 Mm3/d).
In terms of expenses, in the Constrained Investment Scenario CAPEX and OPEX
reached 7.8 billion US$ and 3.5 billion US$ in nominal terms, respectively,
accumulated from 2015 to 2030, whereas in the Unconstrained Investment Scenario,
136
they reached 36 billion US$ (CAPEX) and 11.4 billion US$ (OPEX) for the same
period. It is worth emphasizing that these figures do not reflect all upstream investment
efforts in Argentina, which in the first years were focused on conventional gas (Figure
5.22) but helps to unveil an order of magnitude that is required to tap the
unconventional resources in this country.
Figure 7.1. Unconventional non-associated gas production in Argentina under the
different scenarios assessed
In the Unconstrained investment scenario, there is no restriction on CAPEX for
Argentina from 2016 onwards. Therefore, the unconventional gas produced in this
scenario is the economically unconventional resources feasible to produce until 2030
limited by the market and the production costs.
It is not possible to determine that this is the economic potential of
unconventional resources in Argentina. For this to be done, a larger temporal horizon is
needed. Interestingly enough, the CAPEX investment in the Unconstrained Scenario
stopped in 2029. In 2030 the production is only based on already developed
unconventional fields, this might be the results of a perfect foresight until 2030 as the
137
last year of analysis without planning any capacity to supply natural gas for further
years.
Regarding the market constraints, the natural gas destinations for the
unconventional production were only the domestic Argentinian market and exports
during summer for Chile. The model did not choose to implement a liquefaction export
plant even in the Unconstrained Scenario. This derives from the assumed costs and
LNG export prices.
In terms of costs, the assumed CAPEX for the liquefaction facility was 1,400
US$/tonnes of capacity per year (30.6US$/MMBtu of capacity per year), using a 30
year-life facility, a 10% discount tax and a capacity factor of 100%. The levelized cost
of this facility is 3.2 US$/MMBtu.
Figure 7.2 shows the levelized cost of the Argentinian non-associated gas
production (see also Table 5.6). By summing the levelized CAPEX with OPEX
production for the unconventional resources, a result of 5.3 US$/MMBtu is found.
Consequently, the current export LNG price assumed (5.3 US$/MMBtu) covers the
production costs but fail to cover the liquefaction facility cost.
Figure 7.2. Levelized cost of Argentinian non-associated gas production used in
TIMES-ConoSur.
138
It is relevant to recognize the lack of learning curves for natural gas production
costs in the developed TIMES-ConoSur model. For instance, many unconventional
basins in the USA have increased efficiency in drilling and completion, resulting in an
average cost in 2015 of 25-30% lower than 2012 values (EIA, 2016c). The same would
be expected for Argentina as it develops its unconventional resources. Consequently,
the possibility to expand its market for the unconventional gas by implementing an
LNG exportation facility will depend on the upstream cost reduction achieved by
Argentina in the following years.
Figure 7.3. Average Well Drilling and Completion Cost for some Unconventional
Basins in the USA. Source: EIA (2016c)
7.2 What role LNG will play in the natural gas market of the Southern Cone?
Since the start of operation of Bahia Blanca Gas Port in Argentina, a
regasification capacity of 95.5 Mm3/day has been installed in the Southern Cone.
Historically, LNG terminals were planned to deal with contingencies as a short-term
solution.
In Argentina, Bahia Blanca was implemented under the context of a declining
natural gas production and as an energy security measure to face Bolivian gas supply
disruptions. In Chile, it was more evident the contingency solution to deal with the
139
Argentinian natural gas exports cuts in 2006 and lower the cost of using oil products for
power generation. In Brazil, the fear of Bolivia cutting off exports was repeated in 2006.
However, in spite of increasing the energy security to face Bolivian gas exports possible
cuts, LNG in Brazil play a key role for the drought years (2012-2015) complementing
the hydropower generation.
At this time, LNG increased its competitiveness globally. From all the natural gas
traded globally in 2015, LNG represented 32.5%(BP, 2016). Competition among
exporters has become more drastic and complex, which has been changing from
regional competition to global competition (CHEN et al., 2016; WOOD, 2016).
LNG has become a global commodity, because of the volumes globally traded
and the de-regionalization of LNG (BARNES and BOSWORTH, 2015). However, gas
prices around the world vary widely, and these price differentials have persisted for
years, becoming even more pronounced since the Fukushima accident in 2011 (RITZ,
2014). In terms of contract trade, the LNG markets are moving toward a larger
proportion of volumes being traded on short-term contracts or sold as spot cargoes,
which will promote the liquidity of the LNG market (HARTLEY, 2015;
KHALILPOUR and KARIMI, 2011). All these factors will impact the LNG import
prices in the Southern Cone.
To assess this impact, this study performed LNG price sensitivity scenarios on the
same LNG import prices of the first quarter of 201620
assumed for each country shown
in the Figure 5.21, and the base LNG export prices assumed for all countries (5
US$/MMBtu). The sensitivities simulated were the-3 US$/MMBtu, +5 US$/MMBtu
and +10 US$/MMBtu.
20The LNG import prices estimated for the first quarter fo 2016 were: 6.9 US$/MMBtu in
Argentina, 6.7 US$/MMBTu in Brazil, 4.5 US$/MMBtu for Chile.
140
Figure 7.4. Accumulated imports (+)/exports (-) of LNG from 2016 until 2030 for
different LNG price changes.
Figure 7.4 shows the accumulated international trade of LNG for each country of
the Southern Cone and for each LNG price scenario. The reduction of imports of LNG
when the import price of LNG increases reflects the elasticity in the power generation
sector and the competition with domestic resources. Simultaneously, when LNG prices
rise, the operation of oil products and coal power plants grows and the domestic
production of natural gas is also boosted. The same principle works in the other
direction when LNG prices go down.
Interesting enough, an increase of +5 US$/MMBtu could turn feasible a
liquefaction plant for Brazil even in the Constrained Scenario. The same increase in the
Unconstrained Scenario allows LNG exportation in Argentina and Brazil. In the case of
Chile, only an increase of 10 US$/MMBtu would turn feasible the production of its
unconventional natural gas for export.
This finding raises an important question: for the Southern Cone as a whole, is it
more convenient to have a high or a low LNG international price?
141
Actually, this depends on the investment capacity in the upstream. As seen in
Table 7.1, a reduction of -3 US$/MMBtu would result in costs savings. However, to
these countries to be able to produce their gas resources a scenario of +10US$/MMBtu
becomes crucial, even at higher total system discounted costs. So, the role of LNG in
these countries will depend both on the LNG prices and the ability to produce domestic
production. Besides, LNG regasification and storage facilities will keep their relevance
to maintain a backup source of gas for any contingency such drought years, delays in
the start of new power plants, disruption in the supply of gas, etc.
Table 7.1. Total system discounted cost (MUS$) for the different LNG prices
sensitivities
Sensitivity
Constrained
Scenario
Unconstrained
Scenario
Base 34,653 34,263
-3 33,126 33,532
+5 35,671 34,171
+10 35,982 32,999
7.3 Will Brazil be gas self-sufficient using its associated gas production from
offshore fields?
As seen in Chapter 6 (Figure 6.4), in the Unconstrained Scenario, no LNG imports
are required in Brazil under normal conditions (regular rainfalls and entry of power
plants as planned).
In addition, in this chapter (Figure 7.4), this study has shown that even in a
Constrained Scenario, with 5 US$/MMBtu above of the current prices, LNG imports are
not needed. However, Brazil relies also on the Bolivian gas imports. It is expected to
have a renewal of the gas supply agreement in 2019 (new conditions are currently on
negotiation).
142
As shown in Chapter 6 (Figure 6.3 and Figure 6.4), for both production scenarios,
the Bolivian gas still plays a major role and the differences of volumes imported are
insignificant. However, a compulsory renewal of the GSA contract and a lower limit of
the monthly capacity factor of the pipeline was assumed in the model for both scenarios
in an attempt to emulate the take-or-pay clause.
An additional analysis was made in this chapter, then, for the Unconstrained
Scenario, to test the natural gas self-sufficiency potential in Brazil. For this, from 2019
onwards imports from Bolivia were cut off. The same was done, with LNG, and any
eventual imports from Argentina. This sensitivity scenario is useful to test the offshore
associated gas potential in Brazil.
Figure 7.5 shows the results of these assumptions for Brazil, summed the
Integrated and North Brazilian regions.
Figure 7.5. Natural gas supply in Brazil in a self-sufficient sensitivity scenario.
Brazil only relying on domestic natural gas production was able to attend its internal
demand of gas, becoming a natural gas self-sufficient country domestic from 2019
onwards. Figure 7.6 provides more insights about the sources of domestic production.
As can be observed, to fulfill this scenario, it is necessary to tap new non-associated gas
0
20
40
60
80
100
120
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
MMm3/d
Imports from Bolivia Domestic Production Installed Regasification Plants
143
reserves in Integrated Brazil and in the Amazon basin. In 2030 these provide 21 Mm3/d
and 15 Mm3/d respectively.
Figure 7.6. Gross natural gas production in Brazil in the self-sufficient sensitivity
scenario.
However, this sensitivity of the Unconstrained Scenario represents an additional cost of
0.5 billion U$ at present value. This small difference is explained by the fact that
Brazilian associated gas is a low-cost resource.
Nevertheless, this sensitivity scenario poses two challenges:
1. The first is related to the financial needs. Only for the CAPEX related to natural
gas infrastructure in upstream and midstream, it is required approximately 20
billion US$ from 2019 to 203021
. As Petrobras, which faces financial struggles,
and other players are focusing its expenses on Pre-Salt oil production, additional
incentives will be required to attract other players to tap non-associated gas
reserves in Brazil, such as it is performed in the Mexilhão field.
2. The second challenge is related to the hydrocarbon basins in the Amazonia
region, such as the Solimões Basin, which poses risks for both social conflicts
and environmental impacts in addition to the higher costs of these resources.
21This value does not incorporate the CAPEX related to offshore oil production
0
20
40
60
80
100
120
140
160
2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030
MMm3/d
Amazonian Undeveloped Non-Associated Gas Amazonian Developed Non-Associated Gas
Amazonian Associated Gas Undeveloped Non-Associated Gas Integrated Brazil
Developed Non-Associated Gas Integrated Brazil Pre-Salt Associated Gas
Post-Salt Associated Gas
144
Furthermore, cutting-off Bolivian imports is uneconomical, as the Gasbol pipeline
is an already amortized asset. Besides, to dispense low-cost Bolivian resources results in
losing the opportunity for Brazil to become a net exporter of LNG, if the price rises +5
US$/MMBtu as shown in Figure 7.4. Undeniably, this production potential of domestic
natural gas in Brazil allows his country to better renegotiate with Bolivia the gas supply
agreement terms.
7.4 Will regional natural gas trade in the Southern Cone increase with the
current pipeline infrastructure?
To answer this question, the following regional cases should be checked: exports
from Bolivia to Argentina and Brazil, exports from Argentina to Chile and Brazil.
Regarding Bolivian exports, as observed in chapter 6 (Figure 6.7 and Figure 6.8),
they remain approximately the same in both Constrained and Unconstrained Scenario.
The construction of these scenarios relied on the assumption that both contracts with
PETROBRAS and ENARSA will be renewed in 2019 and 2027, respectively; in
addition, as mentioned before, the take-or-pay clauses were attempted to be
incorporated using lower limits of natural gas imports per year.
In order to test the feasibility of the Bolivian exports take-or-pay clauses, the
lower limits constraints were excluded from 2019 onwards for both Brazil and
Argentina in a sensitivity analysis of the Unconstrained Scenario, which is the most
optimistic for Brazilian and Argentinian domestic production. The trade results between
Bolivia-Brazil and Bolivia-Argentina for the Constrained, Unconstrained and the
sensitivity excluding lower limits in the Unconstrained scenario are shown in Figure 7.7
and Figure 7.8.
145
Figure 7.7. Natural gas trade between Bolivia and Brazil under different scenarios
Figure 7.8. Natural gas trade between Bolivia and Argentina under different
scenarios
The trade results showed in Figure 7.7 and Figure 7.8 do not vary significantly
even excluding the lower limits. This reaffirms that Bolivian exports compose the least-
cost expansion and operation of supply gas in the Southern Cone. This remains valid
event with unconventional gas in Argentina and an optimistic production scenario of
Brazilian offshore oil fields.
25
26
27
28
29
30
31
32
33
34
35
20
12
-M0
12
01
2-M
07
20
13
-M0
12
01
3-M
07
20
14
-M0
12
01
4-M
07
20
15
-M0
12
01
5-M
07
20
16
-M0
12
01
6-M
07
20
17
-M0
12
01
7-M
07
20
18
-M0
12
01
8-M
07
20
19
-M0
12
01
9-M
07
20
20
-M0
12
02
0-M
07
20
21
-M0
12
02
1-M
07
20
22
-M0
12
02
2-M
07
20
23
-M0
12
02
3-M
07
20
24
-M0
12
02
4-M
07
20
25
-M0
12
02
5-M
07
20
26
-M0
12
02
6-M
07
20
27
-M0
12
02
7-M
07
20
28
-M0
12
02
8-M
07
20
29
-M0
12
02
9-M
07
20
30
-M0
12
03
0-M
07
MMm3/d
Unconstrained Unconstrained sensitivity(without lower limits) Constrained
10
12
14
16
18
20
22
24
26
28
20
12
-M0
12
01
2-M
07
20
13
-M0
12
01
3-M
07
20
14
-M0
12
01
4-M
07
20
15
-M0
12
01
5-M
07
20
16
-M0
12
01
6-M
07
20
17
-M0
12
01
7-M
07
20
18
-M0
12
01
8-M
07
20
19
-M0
12
01
9-M
07
20
20
-M0
12
02
0-M
07
20
21
-M0
12
02
1-M
07
20
22
-M0
12
02
2-M
07
20
23
-M0
12
02
3-M
07
20
24
-M0
12
02
4-M
07
20
25
-M0
12
02
5-M
07
20
26
-M0
12
02
6-M
07
20
27
-M0
12
02
7-M
07
20
28
-M0
12
02
8-M
07
20
29
-M0
12
02
9-M
07
20
30
-M0
12
03
0-M
07
MMm3/d
Unconstrained sensitivity(without lower limits) Unconstrained Constrained
146
For the natural gas trade between Argentina and Chile, as mentioned before, in
2016, Argentina signed an agreement with Chile to import natural gas in the winter
season, using the Gas Andes and Norandino pipelines infrastructure and the
regasification capacity of GNL Quinteros and Mejillones plants. The agreement is for 3
Mm3/day through Gas Andes pipeline, including the possibility to increase it by 1
Mm3/day, and 1.5Mm3/day through Norandino pipeline.
These transfers already started in the winter of 2016. Bi-directional flows were
not modelled in the results shown in Chapter 6. However, a sensitivity analysis is
proposed in this section to test a more accurate least-cost trade for Argentina and Chile.
For modelling purposes from 2017 on, the transfers are not limited by the commitments
defined in the Gas Andes and Norandino pipelines agreements, but by the nominal
capacity of these pipelines.
Figure 7.9 shows the results for the trade between Argentina and Chile
incorporating bi-directional flows in Gas Andes and Norandino pipelines.
147
Figure 7.9. Natural gas trade between Argentina and Chile under different
scenarios.
Natural gas flows from Chile to Argentina are observed to be meaningful from
2016 to 2020 in the Constrained Scenario and only in 2016 in the Unconstrained
Scenario. Therefore, the model suggests that with the current LNG selling prices, if new
regasification plants are installed in this decade, the gas transfer from Chile to Argentina
can become a short-term solution.
On the other hand, in the Unconstrained Scenario, exports from Argentina occur
in the summer season at full capacity of the pipelines to Central-South Chile and a
partial load to North Chile.
148
Finally, regarding Argentina-Brazil trade, which corresponds basically to the
supply of natural gas to Brazil’s Uruguaiana combined cycle power plant, Figure 7.10
shows the exports through the Aldeia Brasileña-Uruguaiana pipeline between 2016 and
2030. As observed, gas transfers happened only under the Unconstrained Scenario, and,
as expected, these exports occur in drought months of Brazil (August to December).
Figure 7.10. Natural gas exports from Argentina to Brazil under different
scenarios.
In general, Argentinian unconventional production of natural gas is pivotal to
increase the regional natural gas trade in the Southern Cone. Differences from the
original scenarios show that between 2016 and 2030, in the Unconstrained Scenario the
accumulated trade in the Southern Cone is 2 Tcf higher than the Constrained Scenario
being mostly explained by Argentinian exports.
Finally, the TIMES-ConoSur Model did not consider the modelling of Uruguay.
Therefore, this study omitted the possible natural gas trade between Argentina and
Uruguay trough the “Litoral” pipeline, which was planned to supply the Paysandú city
and the “Cruz del Sur” pipeline, serving the south region of Uruguay, including
Montevideo. Actually, in May of 2016 Argentina and Uruguay subscribed an agreement
to allow gas exports from Uruguay to Argentina using the overcapacity of the “El Plata”
LNG regasification plant projected to be installed in 2017(MIEM, 2016). The modelling
0
0.5
1
1.5
2
2.5
20
16
-M0
1
20
16
-M0
7
20
17
-M0
1
20
17
-M0
7
20
18
-M0
1
20
18
-M0
7
20
19
-M0
1
20
19
-M0
7
20
20
-M0
1
20
20
-M0
7
20
21
-M0
1
20
21
-M0
7
20
22
-M0
1
20
22
-M0
7
20
23
-M0
1
20
23
-M0
7
20
24
-M0
1
20
24
-M0
7
20
25
-M0
1
20
25
-M0
7
20
26
-M0
1
20
26
-M0
7
20
27
-M0
1
20
27
-M0
7
20
28
-M0
1
20
28
-M0
7
20
29
-M0
1
20
29
-M0
7
20
30
-M0
1
20
30
-M0
7
MMm3/d
Unconstrained Constrained
149
of Uruguay is not within the scope of this thesis. However, further studies should
consider this country to better depict the natural gas markets of the Southern Cone.
7.5 Will renewable energy affect the natural gas consumption in power
generation?
To answer this question two topics are addressed: the hydropower expansion in
Brazil and the increase of non-renewable energy.
Hydropower capacity expansion is considered in each region assessed in this
study. Nevertheless, the hydro projects planned in North Brazil is remarkable by both
the total capacity and for the pronounced seasonality.
There are many risks for the energy system when hydropower plants are planned.
One is the entry date of operation and the actual realization of the project. For
hydropower plants planned in the Amazonia, this is critical since these projects face
social opposition by the environmental impacts (ANDRADE and DOS SANTOS, 2015;
FEARNSIDE, 1999) and also have difficulties in obtaining the environmental license.
According to ANDRADE and DOS SANTOS (2015), there were hydroelectric
project alterations even before the environmental licensing process was filed at the
environmental body or as a consequence of an initial refusal by the federal licensing
institution to issue the respective license. The process is also slow, taking around 5
years for the federal agency to give the final answer to the prior environmental license
request (ANDRADE and DOS SANTOS, 2015).
To illustrate this risk, in 2016 the environmental license for the São Luiz do
Tapajós hydropower already planned in EPE (2015b) to start operation in 2023 was not
granted due to non-solved displacement impacts on domestic communities (IBAMA,
2016).
150
Another risk factor for the operation and planning of hydropower in the energy
system is the dependence on rainfall. The seasonality of rain is more pronounced in the
North Brazil region and the drought years pose a risk to the operation of the energy
system.
Both the delays and droughts result in an increase of the natural gas-fired power
plants and consequently in the gas consumption estimated for Brazil. To assess this fact,
a sensitivity analysis was simulated in the Constrained Scenario where the São Luiz dos
Tapajos is cancelled and all hydropower plants in both North Brazil and Integrated
Brazil regions have a capacity factor similar to the dry year of 2015.
Figure 7.11.Natural gas supply sensitivity scenario for Brazil excluding São Luiz
de Tapajos hydroelectric project and with drought years.
As can be observed in Figure 7.11, excluding the São Luiz de Tapajos
hydropower complex and having drought years boosts LNG imports from 2017
onwards. This includes also the necessity of a new regasification plant in 2022.
This highlights the energy security feature of LNG projects in Brazil. Interestingly
enough, the cost difference between this sensitivity analysis and the base scenario
results only in a total discounted net cost of 181 MUS$ at present value. Prudence might
0
50
100
150
200
250
20
12
-M0
12
01
2-M
07
20
13
-M0
12
01
3-M
07
20
14
-M0
12
01
4-M
07
20
15
-M0
12
01
5-M
07
20
16
-M0
12
01
6-M
07
20
17
-M0
12
01
7-M
07
20
18
-M0
12
01
8-M
07
20
19
-M0
12
01
9-M
07
20
20
-M0
12
02
0-M
07
20
21
-M0
12
02
1-M
07
20
22
-M0
12
02
2-M
07
20
23
-M0
12
02
3-M
07
20
24
-M0
12
02
4-M
07
20
25
-M0
12
02
5-M
07
20
26
-M0
12
02
6-M
07
20
27
-M0
12
02
7-M
07
20
28
-M0
12
02
8-M
07
20
29
-M0
12
02
9-M
07
20
30
-M0
12
03
0-M
07
MMm3/d
Imports from Argentina New Regasification Plants Installed Regasification Plants
Amazonian Production Domestic Production Cuiaba Power Plant Imports
Imports from Bolivia
151
imply in investing in a new regasification plant in Brazil to start operations at the
beginning of the next decade.
Moreover, this scenario does not incorporate climate change impacts in
hydropower generation. According to LUCENA et al. (2009), under climate change
scenarios, hydropower in Brazil might lose 3.15% of its firm power. A loss of 2.5 % of
energy produced is projected by the authors in the Paraná Basin and 7% for the São
Francisco Basin. According to them, this loss of hydropower energy would be covered
by natural gas power plants.
Regarding the wind and solar power, the model adopted capacity expansions of
other studies, including governmental official plans. According to these scenarios, non-
conventional renewable increases their capacity share in the energy mix.
Figure 7.12 shows the electricity produced by sources in the Southern Cone in
2015 and 2030, according to the Constrained Scenario. Wind and solar increase whereas
natural gas reduce its market share.
Figure 7.12. Power generation simulated by sources in the Southern Cone in 2015
and 2030
Wind and solar technologies costs have been increasing at a fast pace in the last
years. For instance, in 2016 in a power auction in Chile, the lowest prices for the wind
and solar bids reached 38.0 US$/MWh and 29.1 US$/MWh respectively (CNE, 2016d).
152
Still, this study indicates that a high penetration of renewables in the energy mix
requires natural gas to complement them in the operation. This can be illustrated in a
country like Chile, which will have a high penetration of solar technologies in 2030
(Figure 7.13).
Figure 7.13. Power operation in Chile simulated for 2030.
As Figure 7.13 shows, natural gas combined cycle plants are the marginal
technology. In some months, gas-fired power plants operate during the day. However,
their output is incremented during the night, complementing solar. It is important to
highlight that the TIMES-ConoSur model did not incorporate flexibility features.
Consequently, the energy produced by natural gas might be underestimated. In any case,
it can be stated that wind and solar power technologies are likely to reduce the share of
natural gas in the Southern Cone; nevertheless, these technologies will need natural gas
plants as a flexible backup capacity to support their intermittency and stochasticity.
In sum, the results for hydro, wind and solar power shows that the average output
of gas-fired power plants may decrease due to renewables, but LNG regasification
plants and back-up gas-fired power facilities will increase importance to accommodate
climate, meteorological and projects´ risks. This poses the challenges on how to
remunerate these gas facilities.
0
50
100
150
200
250
300
350
400
450
20
30
-M0
1h
01
20
30
-M0
1h
09
20
30
-M0
1h
17
20
30
-M0
2h
01
20
30
-M0
2h
09
20
30
-M0
2h
17
20
30
-M0
3h
01
20
30
-M0
3h
09
20
30
-M0
3h
17
20
30
-M0
4h
01
20
30
-M0
4h
09
20
30
-M0
4h
17
20
30
-M0
5h
01
20
30
-M0
5h
09
20
30
-M0
5h
17
20
30
-M0
6h
01
20
30
-M0
6h
09
20
30
-M0
6h
17
20
30
-M0
7h
01
20
30
-M0
7h
09
20
30
-M0
7h
17
20
30
-M0
8h
01
20
30
-M0
8h
09
20
30
-M0
8h
17
20
30
-M0
9h
01
20
30
-M0
9h
09
20
30
-M0
9h
17
20
30
-M1
0h
01
20
30
-M1
0h
09
20
30
-M1
0h
17
20
30
-M1
1h
01
20
30
-M1
1h
09
20
30
-M1
1h
17
20
30
-M1
2h
01
20
30
-M1
2h
09
20
30
-M1
2h
17
GWh/(h-month)
Geothermal Hydro Biomass Coal Wind Solar CC Natural Gas
153
The following Table summarizes the research questions on the key factors that
will determine the future natural gas dynamics in the Southern Cone and the answers
that can be stated based on the quantitative insights provided by TIMES-ConoSur.
Table 7.2. Research questions and answers provided by TIMES-ConoSur
Research Question Answers What is the economic potential of
unconventional resources in Argentina?
The economic potential of unconventional gas in Argentina
goes beyond 2030, but with current LNG prices is limited by
the domestic market and exports to pipeline-interconnected
countries. In 2030, without financial restrictions, it is
estimated a production of unconventional gas resources
around 85 Mm3/d.
What role LNG will play in the natural
gas market of the Southern Cone?
The role of LNG in the Southern Cone will depend on the
capacity of the countries to produce economic domestic gas
resources. In a conservative domestic production scenario,
LNG imports are projected to increase in all countries with
access to sea whereas in an optimistic domestic production
scenario LNG is projected to increase only in Chile, however
it will be fundamental to balance the energy system in case
of droughts or any contingency. With an increase of +5
US$/MMBtu of LNG prices, LNG exports projects turn
feasible in Argentina and Brazil.
Will Brazil be gas self-sufficient using
its associated gas production from
offshore fields?
Brazil has the potential to be gas self-sufficient tapping the
associated gas produced in offshore oil fields and non-
associated gas from the Amazon. However, self-sufficiency
is not least cost solution. Keeping imports from Bolivia
provides a better economy and is key for Brazil if the
country decides to implement LNG export projects.
Will natural gas trade in the Southern
Cone increase with the current pipeline
infrastructure?
Regional natural gas trade is likely to increase in the
Southern Cone. This happens not only by the increase of
Bolivian exports to Argentina but because Argentinian can
retake to Brazil and Chile if unconventional resources are
successfully produced.
Will renewable energy affect the natural
gas consumption in power generation?
The average output of gas fired-power plants will decrease in
the Southern Cone due to the planned wind, solar and hydro
power plants. However, gas-fired plants will be pivotal to
complement and backup the high penetration of stochastic
and intermittent renewable technologies. Capacity and
flexibility remuneration mechanisms for gas-fired power
plants are needed to be explored.
154
8 FINAL REMARKS
This thesis sought to assess the pivotal factors that will determine the future
natural gas markets dynamics in the Southern Cone: Argentina, Bolivia, Brazil, and
Chile. The factors identified and analysed were: the unconventional natural gas in
Argentina, LNG imports, the expansion of renewable power technologies, the
associated gas from offshore fields in Brazil and the regional trade of natural gas by
existing pipelines.
There is a lack of literature addressing these issues in Latin America. Most of the
academic studies assessed the energy system modelling in Brazil but with no special
focus on natural gas. In the world, there are natural gas studies analysing some of the
regional factors identified in Europe. This thesis contributes with the academic literature
on natural gas modelling, energy planning and regional energy studies in Latin
America.
For this purpose, a modelling approach was applied. The TIMES platform was
used to model the natural gas markets assuming perfect competition and perfect
foresight. The model built in this thesis, TIMES-ConoSur, incorporated upstream and
midstream process, power generation technologies and the economics of the natural gas
supply chain. TIMES-ConoSur was the main tool to project the supply and demand of
natural gas in the Southern Cone until 2030 using a monthly time resolution and a 6-
region spatial resolution. The outputs of this model indicate the expansion and operation
of the natural gas infrastructure under the least cost performance for the Southern Cone,
as a whole. The objective sought with TIMES-ConoSur was to provide quantitative
insights to understand the role played by key factors in the Southern Cone natural gas
markets.
The critical variable chosen to distinguish two main scenarios in TIMES-ConoSur
was the CAPEX in upstream production. This was an attempt to reflect the financial
constraints to perform economically investments for tapping indigenous resources.
Hence, two scenarios were projected: the Constrained investment scenario, where
155
Argentinian investments for both conventional and unconventional gas production were
limited and a lower offshore oil production projection for Brazil was adopted; and the
Unconstrained investment scenario, with no financial limits for upstream in Argentina
and with an optimistic offshore oil production projection for Brazil. In addition,
sensitivities scenarios were simulated to provide additional insights on each of the
specific key factors assessed.
These two scenarios depicted the interplay in the natural gas supply showing the
competition between natural gas resources from different regions, LNG imports, and
regional trade. Also, on the demand side, the scenarios showed how the availability and
the cost of natural gas affect the power sector operation and the natural gas
consumption.
In the Constrained Scenario the gross domestic production of the Southern Cone
accumulated from 2012 to 2030 was 62 Tcf whereas in the Unconstrained Scenario it
rose to 75 Tcf for the same period. This illustrates the effect of the economic potential
of the unconventional gas resources of Argentina and an optimistic projection of
associated gas from the Campos and Santos basins in Brazil.
Unconventional gas in Argentina can be a game-changer for the natural gas
dynamics in the Southern Cone. According to the model outputs, no financial
restrictions to produce these unconventional resources allowed to reduce projected
accumulated LNG imports in Argentina between 2012 and 2030 from 6.7 Tcf
(Constrained Scenario) to 2.4 Tcf (Unconstrained Scenario). Unconventional gas in
Argentina also allowed this country to retake exports to Chile and Brazil. The results of
the model showed that LNG export prices of 10 US$/MMBtu allowed expanding the
market for Argentinian unconventional resources through the implementation of a
liquefaction onshore facility with potential accumulated LNG exports of 6.3 Tcf until
2030. However, tapping these resources poses a financial challenge. For instance, to
develop the Unconstrained Scenario the investments estimated in upstream for
Argentina were 36 billion US$ (CAPEX) and 11.4 billion US$ (OPEX). There are many
uncertainties about the costs assumed for unconventional gas in Argentina, on the
upwards side cost might rise along sweet spots are depleted first, on the downward side
this study has not considered learning curve effects. Neither this study has
156
disaggregated costs between tight and shale gas. Currently tight gas explains most of
unconventional gas productions. To have a more reliable and detailed upstream cost
data will be key for the robustness of the domestic gas supply projections. In any case,
the government of Argentina should keep an economic stability environment and put
efforts to design an attractive regulatory framework.
LNG imports will remain the balance source of natural gas supply in Brazil and
Argentina. The model showed a strong competition with domestic production for these
two countries. Accumulated LNG imports from 2012 to 2030 in the Southern more than
doubled in the Constrained Scenario (13 Tcf) when compared to the Unconstrained
Scenario (6 Tcf). However, even in the Unconstrained LNG still remains as the most
economic option source to supply peak demands in winter for Argentina and Chile and
in the Constrained scenario LNG will play a major role in Brazil to feed gas-fired power
plants during the dry-season (June to December). LNG imports in Chile were projected
to increase more than threefold the import of 2012 at the end of 2030 for both scenarios.
Nevertheless, these results do not incorporate severe climate conditions such as El Niño
that will upward the volumes of LNG imports during the dry-years.
The sensitivity scenario of LNG prices showed that demand for LNG is highly
elastic. For instance, in the Constrained Scenario with LNG prices of the first quarter of
2016, accumulated LNG imports between 2016 and 2030 were 11.5 Tcf whereas with -3
US$/MMBtu it raised to 19.5 Tcf and with +10US$/MMBtu it declined to 3.3 Tcf.
Brazilian associated gas from offshore oil fields was pivotal too. Although it did
not cause any breakthrough effects in the other countries’ natural gas markets, it had a
dramatic effect on LNG imports, almost eliminating them in Brazil from 2017 up to
2030 in an optimistic oil production scenario. Much of the associated gas production in
Brazil was re-injected, as there was no market for it. However, there are uncertainties
measures to reduce Bolivian imports to open market for this associated gas and also
about the costs of its production which involves different CAPEX costs for fields
according to factors such as CO2 content, existing transport infrastructure or distance to
the coast.
157
This study also tested if associated gas can contribute to making Brazil a self-
sufficient natural gas country. In this case, additional non-associated gas production
from offshore fields and from the Amazon (the Solimões basin, for instance) were
required. Nevertheless, this was not the most economic option as it is cheaper to
continue importing gas from Bolivia. In addition, with an increase of 5 US$/MMBtu in
LNG prices, Brazil’s LNG exports became feasible through a liquefaction plant, in both
Constrained and Unconstrained scenarios. An adequate regulatory framework and
design of the natural gas market have to be explored to establish a proper environment
to attract other players besides of Petrobras to perform the investments for this
infrastructure. In addition, other more modelling tools with a higher granularity on the
upstream and able to incorporate logistics constraints in the gas pipeline network should
be more appropriate to address this case. Also, there are uncertainties about oil and gas
prices evolution in the future; these are key exogenous factors that also will drive the
production of associated gas.
Regional natural gas trade is likely to increase in the Southern Cone. In the
Unconstrained scenario Argentina gas retook exports to Chile during the summer season
using the spare capacity of the non-associated gas fields. In this scenario, Argentina also
sent gas to the Uruguaiana combined cycle plant in Brazil during the dry-season.
Bolivian exports are projected to increase in Argentina and maintain the same values in
Brazil for both scenarios. Bolivia gas trade with Argentina and Brazil remains
approximately the same between the scenarios, suggesting that keeping importing from
Bolivia is the most economic option. This is an outcome based on the logic of regional
costs minimization of the Southern Cone, the modelling only included costs of
production and transport but not the real delivered price. Therefore, an approach
considering costs minimization for either only Brazil or only Argentina, and assuming
gas delivered prices, would likely result in lower gas imports from Bolivia than those
projected in this study.
Also, the regional flows obtained in this study only accounted for the existing gas
pipeline infrastructure, new international pipelines and including Uruguay in the model
might change these outcomes.
158
Regarding the impacts of renewables on natural gas, it can be concluded that wind
and solar energy will undermine the share of natural gas-fired power plants in the
electricity produced in the Southern Cone, reducing it from 22.4% in 2012 to 14.8% in
2030. However, natural gas power plants will increase its relevance as back-up power
facilities to deal with the intermittence and stochasticity of the renewable energy
sources. This aspect is critical in Brazil, where hydropower and wind are projected to
increase significantly. During dry years gas-fired power plants and LNG imports will be
key to have a reliable electricity supply. Capacity mechanism incorporating flexibility
features for these facilities shall be designed to remunerate these power plants and the
availability of the fuel.
As highlighted these are two arbitrary scenarios, where the outcomes for the
natural gas markets are consequence from choices on the assumptions. The conclusions
that can be extracted from this modelling exercise are, therefore, biased in this
framework. A set of alternative hypothesis are strongly suggested to be tested in further
studies to build a more insightful and robust perspective on the Southern Cone gas
markets.
There are three unique aspects of modelling exercise performed in this thesis: it is
the first academic regional natural gas model for energy planning in Latin America; it
involved the developing of a methodology to project natural gas production to deal with
the lack of disaggregated data; and it incorporates some particular features such as
condensates and natural gas liquids productions and its effect on the shadow price of
dry natural gas, field production curves, financial constraints, gas reinjection, gas-to-
liquids, floating LNG facilities, etc. The academic contribution of this thesis is richer on
the methodology proposed to model the natural gas chain in the Southern Cone than the
results provided in the two scenarios simulated.
Nevertheless, there are many features in the natural gas and energy markets that
TIMES-ConoSur and this study did not address. One of them is related to the spatial
scope of this study, omitting Uruguay and Paraguay. This might affect the natural gas
trade results in Argentina both by the new GNL del Plata terminal and the potential for
gas production in offshore basins in the case of Uruguay, and the potential market for
gas and the electricity exports in the case of Paraguay. The spatial resolution also
159
adopted for this study did not allow capturing the congestion in domestic pipelines. This
might become relevant in the south region of Brazil in the TBG pipelines, for instance.
Also regarding pipelines, the results were only based on existing international
pipelines. New pipelines can expand the markets and potentially reduce the costs in the
Southern Cone. Existing natural gas storage in Argentina was also omitted; its inclusion
could have reduced LNG imports estimated during the winter season.
On the upstream side, Government take is not commonly incorporated in energy
planning models and TIMES-ConoSur did not incorporate it either. However, taking
into account this factor might divert investments to produce natural gas towards other
countries with higher resources costs but lower government take. Also in upstream,
costs were maintained constant in spite of the fact that learning curve costs might have a
significant impact on natural gas costs production and facilities.
On the demand side, the assumption of inelasticity for natural gas end-uses is not
reasonable in the long-term; however, the currently available data on historical natural
gas prices and consumption in the Southern Cone makes difficult to estimate this
elasticity. An alternative could be modelling end-uses sectors at the level of useful
energy, but available data is also a limitation in this case.
Commercial issues are not captured in this model either. Specifications such as
pricing and take commitments in power purchase agreements and gas supply
agreements, or commercial strategies of companies if considered should impact
significantly on the results of this study. A limitation to include these factors in the
modelling is the limited access to the contracts; however, high-resourced commercial
and governmental models should make efforts to incorporate these issues to improve the
market accuracy of the results.
In terms of the rationale of the model, the main two assumptions of perfect
competition and perfect foresight ignore the real dynamic of the energy markets and
especially the natural resources economics. Myopic foresight simulations are possible to
perform in TIMES. However to represent the market power of the countries, for
instance, the market power of Bolivia as a regional gas supplier, other modelling
160
platforms than TIMES should have been chosen. This is also valid to represent energy
strategies related to geopolitics, such as the political integration of Latin America.
Fair enough to say is that there is a trade-off between including this additional
modelling features not addressed by TIMES-ConoSur and the computational resources
and the execution time required solving the model, especially when a higher
geographical and temporal resolution are implemented.
A possible future key factor for the Southern Cone natural gas market was not
included in this study: climate change mitigation targets. However, a carbon price
methodology is not sufficient to incorporate this factor in TIMES-ConoSur, since the
representation of other energy chains is needed. In other words, an integrated energy
system modelling is needed to assess the interplay between gas and low-carbon energy
resources and technologies.
Finally, further research is required to improve the modelling of natural gas
markets dynamics and overcome the limitations of this study. This includes: the analysis
of the role of natural gas storage in Argentina and future storage infrastructure in Brazil;
the exploring of uncertainties on the assumptions made in this study, the study of how
the gas business models will be remunerated, the evaluation of how climate change
mitigation policies can affect the future natural gas supply and demand; the analysis of
the impacts of the scenarios developed in this study on the whole economy of the
assessed countries; and, the assessment of opportunities to integrate the Southern Cone
with other natural gas markets in Latin America.
161
9 REFERÊNCIAS BIBLIOGRÁFICAS
ADEBOYE, Y. B.; UBANI, C. E.; ORIBAYO, O. Prediction of Reservoir Performance
in Multi-Well Systems Using Modified Hyperbolic Model. Journal of Petroleum
Exploration and Production Technology, v. 1, n. 2–4, p. 81–87, 15 set. 2011.
AGEERA. Proyecto “Escenarios Energéticos Argentina 2030”. Buenos Aires:
Asosiación de Generadores de Energía Eléctrica de la República Argentina, 2012. .
<http://www.escenariosenergeticos.org/fotos/downloads/2012/11/Informe-AGEERA-
Ampliado.pdf>. Accessed on: 8 set. 2015.
AGUILERA, R. F. The role of natural gas in a low carbon Asia Pacific. Applied
Energy, v. 113, p. 1795–1800, jan. 2014.
ALMEIDA, E. L. F. D.; OLIVEIRA, A. D. Developing the gas industry in Brazil:
Competition or regulation? Minerals & Energy - Raw Materials Report, v. 15, n. 3,
p. 14–21, 1 jan. 2000.
ALMEIDA PRADO JR., F.; ATHAYDE, S.; MOSSA, J.; BOHLMAN, S.; LEITE, F.;
OLIVER-SMITH, A. How much is enough? An integrated examination of energy
security, economic growth and climate change related to hydropower expansion in
Brazil. Renewable and Sustainable Energy Reviews, v. 53, p. 1132–1136, jan. 2016.
ALOSHBAN, G.; SANDREA, I. Global E&P capital expenditure: trends &
determinant. Oil and Gas Financial Journal, 1 fev. 2006.
<http://www.ogfj.com/articles/print/volume-3/issue-2/features/global-eampp-capital-
expenditure-trends-amp-determinants.html>. Accessed on: 25 jan. 2016.
AMORIM, F.; PINA, A.; GERBELOVÁ, H.; PEREIRA DA SILVA, P.;
VASCONCELOS, J.; MARTINS, V. Electricity decarbonisation pathways for 2050 in
Portugal: A TIMES (The Integrated MARKAL-EFOM System) based approach in
closed versus open systems modelling. Energy, v. 69, p. 104–112, Maio 2014.
ANDRADE, A. de L.; DOS SANTOS, M. A. Hydroelectric plants environmental
viability: Strategic environmental assessment application in Brazil. Renewable and
Sustainable Energy Reviews, v. 52, p. 1413–1423, Dezembro 2015.
ANDREWS, G. E. 16 - Ultra-low nitrogen oxides (NOx) emissions combustion in gas
turbine systems A2 - Jansohn, Peter. In: Modern Gas Turbine Systems. Woodhead
Publishing Series in Energy. [s.l.] Woodhead Publishing, 2013. p. 715–790.
ANH. Anuario Estadístico 2014. La Paz, Bolivia: Agencia Nacional de Hidrocarburos,
2015. . <http://descargas.anh.gob.bo/Anh/Anuario2014.pdf>. Accessed on: 8 set. 2015.
162
ANP. Regulamento técnico de queimas e perdas de petróleo e gás natural. 2000.
ANP. Anuário estatístico brasileiro do petróleo, gás natural e biocombustíveis
2007. Rio de Janeiro: Agência Nacional de Petróleo, Gás Natural e Biocombustíveis,
2007. . <http://www.anp.gov.br/?dw=1090>.
ANP. Anuário estatístico brasileiro do petróleo, gás natural e biocombustíveis
2015. Rio de Janeiro: Agência Nacional de Petróleo, Gás Natural e Biocombustíveis,
2015. . <http://www.anp.gov.br/?dw=78135>.
ANP. Dados Estatísticos Mensais.
<http://www.anp.gov.br/?pg=69299&m=&t1=&t2=&t3=&t4=&ar=&ps=&1462366350
864>. Accessed on: 5 abr. 2016a.
ANP. Publicação do Boletím Anual de Reservas (31/12/2015). Rio de Janeiro:
Agência Nacional de Petróleo, Gás Natural e Biocombustíveis, 2016b. .
<http://www.anp.gov.br/?dw=79951>.
ARPS, J. J. Analysis of Decline Curves. Trans. AIME, v. 160, p. 228–247, 1945.
ARTECONI, A.; BRANDONI, C.; EVANGELISTA, D.; POLONARA, F. Life-cycle
greenhouse gas analysis of LNG as a heavy vehicle fuel in Europe. Applied Energy, v.
87, n. 6, p. 2005–2013, jun. 2010.
BARNES, R.; BOSWORTH, R. LNG is linking regional natural gas markets: Evidence
from the gravity model. Energy Economics, v. 47, p. 11–17, jan. 2015.
BCRA. Estadísticas- Principales Variables - Tipo de cambio de referencia (pesos
por dólar estadounidense). <http://www.bcra.gov.ar/Estadisticas/estprv010001.asp>.
Accessed on: 17 maio. 2016.
BHRAN, A. A. E.-K.; HASSANEAN, M. H.; HELAL, M. G. Maximization of natural
gas liquids production from an existing gas plant. Egyptian Journal of Petroleum,
2016. <http://www.sciencedirect.com/science/article/pii/S1110062115300477>.
Accessed on: 3 mar. 2016.
BLACK&VEATCH. Cost and Performance Data for Power Generation
Technologies. [s.l.] Prepared for National Renewable Energy Laboratory, fev. 2012. .
BLS. Producer Price Indexes. <http://www.bls.gov/ppi/>. Accessed on: 27 fev. 2016.
BNEF. Climate Scope 2015. UK: Bloomberg New Energy Finance, 2016. .
<http://global-climatescope.org/en/download/>. Accessed on: 19 ago. 2016.
BOOTS, M. G.; RIJKERS, F. A. M.; HOBBS, B. F. Trading in the Downstream
European Gas Market: A Successive Oligopoly Approach. The Energy Journal, v. 25,
n. 3, p. 73–102, 2004.
BÖRJESON, L.; HÖJER, M.; DREBORG, K.-H.; EKVALL, T.; FINNVEDEN, G.
Scenario types and techniques: Towards a user’s guide. Futures, v. 38, n. 7, p. 723–
739, Setembro 2006.
163
BP. Statistical Review of World Energy 2015. London, UK: British Petroleum, 2016.
. <http://www.bp.com/content/dam/bp/pdf/Energy-economics/statistical-review-
2015/bp-statistical-review-of-world-energy-2015-full-report.pdf>.
BRAGA, Y. C. P. Análise da demanda de gás natural no brasil: uma perspectiva
metodológica. 2014. Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil,
2014. <http://www.ie.ufrj.br/images/pos-graducao/ppge/Yanna_Clara_VF.pdf>.
Accessed on: 15 ago. 2016.
BRANDT, A. R. Review of mathematical models of future oil supply: Historical
overview and synthesizing critique. Energy, v. 35, n. 9, p. 3958–3974, Setembro 2010.
BRUCKNER, T.; BASHMAKOV, Y.; MULUGETTA, H.; CHUM, A.; DE LA VEGA
NAVARRO, J.; EDMONDS, A.; FAAIJ, B.; FUNGTAMMASAN, A.; GARG, E.;
HERTWICH, D.; HONNERY, D. Energy Systems. Climate Change 2014:
Mitigation of Climate Change. Contribution of Working Group III to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change. New York,
USA: International Panel of Climate Change, 2014. .
<https://www.ipcc.ch/pdf/assessment-report/ar5/wg3/ipcc_wg3_ar5_chapter7.pdf>.
CABANES, J. GNL Mejillones, hub en Latam. In: SEMINARIO ARPEL 50
“DESARROLLO DE MERCADOS DE GAS NATURAL”. Santiago, Chile, 2015. .
<http://www.gnlm.cl/sitio/wp-content/uploads/2015/10/Arpel.pdf>. Accessed on: 3 mar.
2016.
CAF/CIER. Nuevas Oportunidades de Interconexión Eléctrica en América Latina:
Infraestructura. Colombia: Corporación Andina de Fomento, 2012. .
<http://publicaciones.caf.com/media/18406/oportunidades-interconexion-electrica-
america-latina.pdf>. Accessed on: 25 jan. 2016.
CAMARGO, T. R. M. de; MERSCHMANN, P. R. de C.; ARROYO, E. V.; SZKLO, A.
Major challenges for developing unconventional gas in Brazil – Will water resources
impede the development of the Country׳s industry? Resources Policy, v. 41, p. 60–71,
Setembro 2014.
CAMAROTA, P. Novos Contratos de Operação Bolivia –Perspectivas de
Desenvolvimento. In: COMISSÃO DE GAS NATURAL. Rio de Janeiro, 27 ago. 2007.
.
CAMMESA. Informe Anual 2014. Buenos Aires, Argentina.: Compañía
Administradora del Mercado Mayorista Eléctrico S.A., 2015. .
<http://portalweb.cammesa.com/memnet1/Pages/descargas.aspx>. Accessed on: 11
maio. 2015.
CAMMESA. Base_informe_mensual_2016-06.
<http://portalweb.cammesa.com/memnet1/Pages/descargas.aspx>. Accessed on: 9 out.
2016.
CAPROS, P.; PAROUSSOS, L.; FRAGKOS, P.; TSANI, S.; BOITIER, B.; WAGNER,
F.; BUSCH, S.; RESCH, G.; BLESL, M.; BOLLEN, J. European decarbonisation
pathways under alternative technological and policy choices: A multi-model analysis.
164
Energy Strategy Reviews, Sustainable Energy System Changes. v. 2, n. 3–4, p. 231–
245, Fevereiro 2014.
CAPROS, P.; TASIOS, N.; DE VITA, A.; MANTZOS, L.; PAROUSSOS, L. Model-
based analysis of decarbonising the EU economy in the time horizon to 2050. Energy
Strategy Reviews, European Energy System Models. v. 1, n. 2, p. 76–84, Setembro
2012a.
CAPROS, P.; TASIOS, N.; DE VITA, A.; MANTZOS, L.; PAROUSSOS, L.
Transformations of the energy system in the context of the decarbonisation of the EU
economy in the time horizon to 2050. Energy Strategy Reviews, European Energy
System Models. v. 1, n. 2, p. 85–96, Setembro 2012b.
CASTELO BRANCO, D. A.; SZKLO, A. S.; SCHAEFFER, R. Co2e emissions
abatement costs of reducing natural gas flaring in Brazil by investing in offshore GTL
plants producing premium diesel. Energy, v. 35, n. 1, p. 158–167, jan. 2010.
CEARE-UBA. Infraestructura de integración eléctrica en el Cono Sur: Situación
actual y perspectivas: Integración Energética Argentina-Brasil. Buenos Aires,
Argentina: Instituto Economía Energética - FB, 2004. .
<http://www.ceare.org/materiales/ieargbra/archivos/ieab8.pdf>. Accessed on: 3 mar.
2016.
CFR. Code of Federal Regulations 40 CFR60.111b. <http://www.ecfr.gov/cgi-
bin/text-
idx?SID=9aeac62c60b56b9bd8ac168f03f2778b&mc=true&node=se40.7.60_1111&rgn
=div8>. Accessed on: 9 out. 2016.
CHAVEZ-RODRIGUEZ, M. F.; DIAS, L.; SIMOES, S.; SEIXAS, J.; SZKLO, A.;
LUCENA, A. F. P.; HAWKES, A. Natural Gas Outlook for the Southern Cone:
outcomes from an hourly basis TIMES natural gas & power model. In: International
energy workshop 2016. Cork, Ireland: 1 jun. 2016a.
CHAVEZ-RODRIGUEZ, M. F.; GARAFFA, R.; ANDRADE, G.; CARDENAS, G.;
SZKLO, A.; LUCENA, A. Can Bolivia keep its role as a major natural gas exporter in
South America? Journal of Natural Gas Science and Engineering, 2016b.
<http://www.sciencedirect.com/science/article/pii/S1875510016303912>. Accessed on:
8 jun. 2016.
CHAVEZ-RODRÍGUEZ, M. F.; GARAFFA, R.; ANDRADE, G.; CARDENAS, G.;
SZKLO, A.; LUCENA, A. F. P. Will Bolivia Keep Its Role as a Major Natural Gas
Exporter in South America? Journal of Natural Gas Science and Engineering, 2016.
CHAVEZ-RODRIGUEZ, M. F.; SZKLO, A.; DE LUCENA, A. F. P. Analysis of past
and future oil production in Peru under a Hubbert approach. Energy Policy, v. 77, p.
140–151, fev. 2015.
CHAVEZ-RODRIGUEZ, M. F.; VARELA, D.; RODRIGUES, F.; BUSTOS, J.;
KOBERLE, A.; VASQUEZ, E. Forecasting Natural Gas Markets between Chile and
Argentina under developments of unconventional gas and expansion of LNG
regasification plants. In: Bergen, Norway. In: 39TH International IAEE conference.
Bergen, Norway: 22 jun. 2016c.
165
CHAVEZ-RODRÍGUEZ, M.; GONZALEZ, E. M.; SCHAEFFER, R.; SZKLO, A.;
LUCENA, A. F. P.; ARANGO, S. A.; MARTINEZ, J. J.; CARVAJAL, P.; EGUEZ, A.
Looking to the andes: long-term scenarios of oil supply and demand for peru, colombia
and ecuador. In: The 37th IAEE international conference. New York: 15 jun. 2014.
<http://www.usaee.org/usaee2014/program_concurrent.aspx>.
CHEN, Z.; AN, H.; GAO, X.; LI, H.; HAO, X. Competition pattern of the global
liquefied natural gas (LNG) trade by network analysis. Journal of Natural Gas
Science and Engineering, v. 33, p. 769–776, jul. 2016.
CHEN, Y.; HAO, MINGQIANG. Application of HCZ model to predicting multiple
production peaks. ACTA PETROLEI SINICA, v. 34, n. 4, p. 747–752, 2013.
CHEN, Y.Q. Derivation and application of Weng´s Prediction model. Natural Gas
Industry, v. 16, n. 2, p. 22–26, 1996.
CHYONG, C. K.; HOBBS, B. F. Strategic Eurasian natural gas market model for
energy security and policy analysis: Formulation and application to South Stream.
Energy Economics, v. 44, p. 198–211, jul. 2014.
CIGÁS. 30% da produção em gasoduto no AM fica sem uso e volta para a terra. G1, 10
abr. 2015. <http://g1.globo.com/am/amazonas/noticia/2015/04/30-da-producao-em-
gasoduto-no-am-fica-sem-uso-e-volta-para-terra.html>.
CNDC. Memoria 2012: Resultados de operación del Sistema Interconectado
Nacional. Cochabamba, Bolivia: Comite Nacional de Despacho de Carga, 2013. .
<http://www.cndc.bo/boletines/memorias.php>. Accessed on: 23 maio. 2015.
CNDC. Plan Eléctrico del Estado Plurinacional de Bolívia 2025. La Paz, Bolivia:
Comite Nacional de Despacho de Carga, 2014. .
<http://www.cndc.bo/media/archivos/boletines/peebol2025.pdf>. Accessed on: 23 maio.
2015.
CNDC. Generacion bruta (Mwh) : año 2015.
<http://www.cndc.bo/media/archivos/estadistica_anual/genbruta_2015.htm>. Accessed
on: 29 jan. 2015.
CNE. Fijación de precios de nudo Abril 2015 Sistema Eléctrico.
<http://www.cne.cl/tarificacion/electrica/precio-nudo-corto-plazo/>. Accessed on: 12
dez. 2016.
CNE. Balance Energético. <http://energiaabierta.cne.cl/balance-energetico/>.
Accessed on: 20 jan. 2016a.
CNE. Catálogo de estadísticas: Electricidad.
<http://energiaabierta.cne.cl/electricidad/>. Accessed on: 20 jan. 2016b.
CNE. Catálogo de Estadísticas: Hidrocarburos.
<http://www.cne.cl/estadisticas/hidrocarburo/>.
CNE. Exitosa licitación de suministro para clientes regulados permitirá bajar la
cuenta de luz a hogares y pymes a partir del año 2021.
166
<http://www.cne.cl/prensa/agosto/exitosa-licitacion-de-suministro-para-clientes-
regulados-permitira-bajar-la-cuenta-de-luz-a-hogares-y-pymes-a-partir-del-ano-2021/>.
Accessed on: 12 dez. 2016d.
COLOMER FERRARO, M.; HALLACK, M. The development of the natural gas
transportation network in Brazil: Recent changes to the gas law and its role in co-
ordinating new investments. Energy Policy, Special Section: Past and Prospective
Energy Transitions - Insights from History. v. 50, p. 601–612, nov. 2012.
COMMEND. LEAP: User Guide.
<http://www.energycommunity.org/default.asp?action=41>. Accessed on: 2 mar. 2016.
CORIA, J. Environmental policy, fuel prices and the switching to natural gas in
Santiago, Chile. Ecological Economics, v. 68, n. 11, p. 2877–2884, Setembro 2009.
COSTAIN. NGL Extraction. London, UK: Costain, [s.d.].
<http://costain.com/media/596755/gas-ngl-extraction-2pp-web.pdf>.
CUNHA, R. P. F. O GNL e a integraçao na america do sul. 2010. Universidade
Federal do Rio de Janeiro, Rio de Janeiro, Brazil, 2010.
<http://www.ie.ufrj.br/images/pos-graducao/ppge/Roberto_Pougy_.pdf>. Accessed on:
28 mar. 2016.
DA COSTA FRAGA, C. T.; BORGES, F. A.; BELLOT, C.; BELTRÃO, R.;
ASSAYAG, M. I. Campos Basin - 25 Years of Production and its Contribution to the
Oil Industry. In: Offshore Technology Conference, 2003.
<http://www.onepetro.org/doi/10.4043/15219-MS>. Accessed on: 3 maio. 2016.
D’ALESSANDRO, A. A.; IZURIETA, E. M.; TONELLI, S. M. Decision-making tools
for a LNG regasification plant siting. Journal of Loss Prevention in the Process
Industries, v. 43, p. 255–262, Setembro 2016.
DALY, H. E.; FAIS, B. UK TIMES model overview. London, UK: University College
London, 2014. . <https://www.ucl.ac.uk/energy-models/models/uktm-ucl/uktm-
documentation-overview>. Accessed on: 14 ago. 2016.
DANDEKAR, A. Critical evaluation of empirical gas condensate correlations. Journal
of Natural Gas Science and Engineering, v. 27, Part 1, p. 298–305, nov. 2015.
DARGAY, J.; GATELY, D.; SOMMER, M. Vehicle Ownership and Income Growth,
Worldwide : 1960 - 2030. Energy Journal, v. 28, n. 4, p. 143–170, 2007.
DE MORAES CRUZ, R. O.; ROSA, M. B.; BRANCO, C. C. M.; DE SANT’ANNA
PIZARRO, J. O.; DE SOUZA SILVA, C. T. Lula NE Pilot Project - An Ultra-Deep
Success in the Brazilian Pre-Salt. In: Offshore Technology Conference, 2016.
<http://www.onepetro.org/doi/10.4043/27297-MS>. Accessed on: 7 maio. 2016.
DEVLIN, J.; LI, K.; HIGGINS, P.; FOLEY, A. The importance of gas infrastructure in
power systems with high wind power penetrations. Applied Energy, 2016.
<http://www.sciencedirect.com/science/article/pii/S0306261915013926>. Accessed on:
20 jan. 2016.
167
DI SBROIAVACCA, N. Shale Oil y Shale Gas en Argentina. Estado de situación y
prospectiva. Bariloche, Argentina: Fundación Bariloche, 2013. .
<http://www.fundacionbariloche.org.ar/wp-content/uploads/2014/11/Shale-oil-y-shale-
gas.pdf>.
DICCO, R. Inversiones en el Sector Hidrocarburifero 2003-2011. Buenos Aires:
Centro Latinoamericano de Investigaciones Científicas y Técnicas, 2011. .
<http://www.cienciayenergia.com/Contenido/pdf/010711_rad_arg.pdf>. Accessed on: 8
set. 2015.
DICCO, R. Avances del Plan Energético Nacional 2004-2019. Buenos Aires: Centro
Latinoamericano de Investigaciones Científicas y Técnicas, 2013. .
<http://www.cienciayenergia.com/Contenido/pdf/010711_rad_arg.pdf>. Accessed on: 8
set. 2015.
DIUANA, F. A.; SZKLO, A.; GARAFFA, R. Análise da opção de importação de GNL
pelo Brasil e perspectivas de futuras compras. In: Rio oil & gas expo and conference
2016. Rio de Janeiro, Brazil: IBP, 2016.
DNRPA, D. N. de R. del A. y C. P. Estadística Anual de Parque Activo.
<http://www.dnrpa.gov.ar/portal_dnrpa/boletines_estadisticos.php#.VT6ZpiF_Oko>.
Accessed on: 27 abr. 2015.
DOS SANTOS, S. P.; EUGENIO LEAL, J.; OLIVEIRA, F. The development of a
natural gas transportation logistics management system. Energy Policy, v. 39, n. 9, p.
4774–4784, Setembro 2011.
E3MLAB/ICCS. PRIMES Model 2013-2014: Detailed model description. Athens,
Greece: National Technical University of Athens, 2014. .
<http://bakerinstitute.org/media/files/Research/81966512/the-baker-institute-world-gas-
trade-model-biwgtm.pdf>. Accessed on: 28 mar. 2016.
E-CL. De la interconexión de los sistemas a la integración regional, 2015. .
<http://www.cigre.cl/Bienal2015/wp-content/uploads/2015/12/ECL.pdf>. Accessed on:
25 maio. 2016.
ECONLINK. La Producción Petrolera en Argentina, Breve Recorrido Histórico.
<http://www.econlink.com.ar/petroleo-argentina/historia>. Accessed on: 17 maio. 2016.
ECONOMIDES, M. J.; WOOD, D. A. The state of natural gas. Journal of Natural Gas
Science and Engineering, v. 1, n. 1–2, p. 1–13, jul. 2009.
ECOPETROL. Nos estamos reinventando para competir en el mundo: Estrategia al
2020. Bogota, Colombia: Ecopetrol S.A., 2015. .
<http://www.ecopetrol.com.co/documentos/Estrategia-Inversionistas-ESP-
26May15.pdf>.
ECOPETROL. Ecopetrol: Marco estratégico.
<http://www.ecopetrol.com.co/wps/portal/es/ecopetrol-web/nuestra-empresa/quienes-
somos/acerca-de-ecopetrol/marco-estrategico>.
168
EGGING, R. Benders Decomposition for multi-stage stochastic mixed complementarity
problems – Applied to a global natural gas market model. European Journal of
Operational Research, v. 226, n. 2, p. 341–353, Abril 2013.
EGGING, R.; HOLZ, F.; GABRIEL, S. A. The World Gas Model: A multi-period
mixed complementarity model for the global natural gas market. Energy, v. 35, n. 10, p.
4016–4029, Outubro 2010.
EIA. Natural Gas Processing: The Crucial Link Between Natural Gas Production
and Its Transportation to Market. Washington, DC, USA: Energy Information
Administration, 2006. .
<https://www.eia.gov/pub/oil_gas/natural_gas/feature_articles/2006/ngprocess/ngproces
s.pdf>.
EIA. Model Documentation Report: Natural Gas Transmission and Distribution
Module of the National Energy Modeling. Washington, DC, USA: Energy
Information Administration, 2013a. .
<http://www.eia.gov/forecasts/aeo/nems/documentation/ngtdm/pdf/m062(2013).pdf>.
Accessed on: 28 mar. 2016.
EIA. Technically Recoverable Shale Oil and Shale Gas Resources: An Asseessment
of 137 Shale Formations in 41 Countries Outside the United States. Washington,
USA: U.S. Energy Information Administration, 2013b. .
<http://www.eia.gov/analysis/studies/worldshalegas/pdf/fullreport.pdf>.
EIA. Technically Recoverable Shale Oil and Shale Gas Resources: Other South
America. Washington, DC, USA: Energy Information Administration, 2013c. .
<https://www.eia.gov/analysis/studies/worldshalegas/pdf/Other_South_America_Bolivi
a_Chile_Paraguay_2013.pdf>. Accessed on: 3 mar. 2016.
EIA. Technically Recoverable Shale Oil and Shale Gas Resources: Argentina.
Washington, DC, USA: Energy Information Administration, 2013d. .
<https://www.eia.gov/analysis/studies/worldshalegas/pdf/Argentina_2013.pdf>.
Accessed on: 6 nov. 2015.
EIA. Oil and Gas Supply Module of the National Energy Modeling System: Model
Documentation 2014. Washington, DC, USA: Energy Information Administration,
2014. .
<https://www.eia.gov/forecasts/aeo/nems/documentation/ogsm/pdf/m063(2014).pdf>.
Accessed on: 28 mar. 2016.
EIA. International Energy Statistics: Chile.
<https://www.eia.gov/beta/international/data/browser/#?iso=CHL&c=00000004&ct=0
&ord=CR&cy=2013&v=H&vo=0&so=0&io=0&start=1980&end=2013&vs=INTL.3-1-
CHL-BCF.A&pa=g1q0000000001&f=A&ug=g&tl_type=p&tl_id=3002-A>. Accessed
on: 21 jan. 2016.
EIA. International Energy Outlook 2016. [s.l.] U.S. Energy Information
Administration, 2016a. . <http://www.eia.gov/forecasts/ieo/pdf/0484(2016).pdf>.
EIA. Glossary. <http://www.eia.gov/tools/glossary/index.cfm>. Accessed on: 2 mar.
2016b.
169
EIA. Trens in U.S Oil and Natural Gas Upstream Costs. Washington, DC, USA:
Energy Information Administration, 2016c. .
<https://www.eia.gov/analysis/studies/drilling/pdf/upstream.pdf>.
EKANEM ATTAH, E.; BUCKNALL, R. An analysis of the energy efficiency of LNG
ships powering options using the EEDI. Ocean Engineering, Energy Efficient Ship
Design and Operations. v. 110, Part B, p. 62–74, Dezembro 2015.
ENAP. Directorio de ENAP aprueba Plan Estratégico 2014-2025 con inversiones de
US$ 800 millones anuales a 2020.
<http://www.enap.cl/sala_prensa/noticias_detalle/general/776/directorio-de-enap-
aprueba-plan-estrategico-2014-2025-con-inversiones-de-us-800-millones-anuales-a-
2020>. Accessed on: 12 out. 2015.
ENAP. Estados de situación financiera consolidados intermedios correspondientes
al periodo terminado al 31 de Marzo de 2016. Santiago, Chile: Empresa Nacional de
Petróleo, 2016. . <http://www.enap.cl/pag/574/1215/ano_2015>. Accessed on: 7 jul.
2015.
ENARGAS, E. N. R. del G. N. Datos Operativos de Gas Natural: Licenciatarias de
Distribución. <http://www.enargas.gov.ar/DatosOper/Indice.php>. Accessed on: 27
abr. 2015.
ENARSA. Informe de Gestión 2014. Buenos Aires: Energía Argentina S.A., 2015. .
<http://www.enarsa.com.ar/images/Informe_de_Gestion_2014/idg2014.pdf>. Accessed
on: 8 set. 2015.
ENDE. Firman acuerdo para financiar proyecto “Línea de Transmisión Juana
Azurduy de Padilla” que interconectará Bolivia con Argentina.
<http://www.ende.bo/noticia/noticia/27>. Accessed on: 18 maio. 2016.
EPE. Séries Históricas Completas.
<https://ben.epe.gov.br/BENSeriesCompletas.aspx>. Accessed on: 20 jan. 2016a.
EPE. Plano Decenal de Expansão de Energia 2024. Rio de Janeiro: Empresa de
Pesquisa Energética, 2015b. .
<http://www.epe.gov.br/Estudos/Documents/PDE%202024.pdf>. Accessed on: 25 jan.
2016.
EPE. Balanço Energético Nacional 2015: Ano base 2014. Rio de Janeiro: Empresa de
Pesquisa Energética, 2016a. .
<https://ben.epe.gov.br/downloads/Relatorio_Final_BEN_2015.pdf>. Accessed on: 25
jan. 2016.
EPE. Balanço Energético Nacional 2016: Relatório Síntese- Ano base 2015. Rio de
Janeiro: Empresa de Pesquisa Energética, 2016b. .
<https://ben.epe.gov.br/downloads/S%C3%ADntese%20do%20Relat%C3%B3rio%20F
inal_2016_Web.pdf>. Accessed on: 25 jan. 2016.
EPE, E. de P. E. Plano decenal de expansão da malha de transporte dutoviário –
PEMAT 2022. Rio de Janeiro: Empresa de Pesquisa Energética, 2014. .
170
<http://www.epe.gov.br/PEMAT/Relat%C3%B3rio%20Final%20PEMAT%202022.pdf
>.
EPRINC. Condensate: An EPRINC Primer. Washington, DC, USA: Energy Policy
Research Foundation, 4 fev. 2015. . <http://eprinc.org/2015/02/condensate-eprinc-
primer/>. Accessed on: 2 mar. 2016.
EXCELERATE ENERGY. Bahía Blanca GasPortExcelerate Energy, 2016. .
<http://excelerateenergy.com/project/bahia-blanca-gasport/>. Accessed on: 19 ago.
2016.
FAPESP. Para desenvolver o gás natural. Pesquisa FAPESP, fev. 2016.
<http://revistapesquisa.fapesp.br/wp-content/uploads/2016/02/032-
033_Gas_Natural_240.pdf?24a723>.
FEARNSIDE, P. M. Social Impacts of Brazil’s Tucuruí Dam. Environmental
Management, v. 24, n. 4, p. 483–495, 1999.
FERIOLI, J. Futuro del petróleo y el gas en la Argentina – Un desafío no convencional.
In: Día nacional de la energía. Buenos Aires, Argentina., 27 ago. 2013. .
<http://www.jorgeferioli.com/presentacion-jorge-ferioli.pdf>. Accessed on: 11 maio.
2015.
FILIPPÍN, C.; LARSEN, S. F. Historical Consumption of Heating Natural Gas and
Thermal Monitoring of a Multifamily High-Rise Building in a Temperate/Cold Climate
in Argentina. Buildings, v. 2, n. 4, p. 477–496, 4 dez. 2012.
FILIPPÍN, C.; LARSEN, S. F.; MERCADO, V. Winter energy behaviour in multi-
family block buildings in a temperate-cold climate in Argentina. Renewable and
Sustainable Energy Reviews, v. 15, n. 1, p. 203–219, Enero 2011.
FRANCO, A. Natural gas consumption and correlation with the uses of thermal energy:
Analysis of the Italian case. Journal of Natural Gas Science and Engineering, v. 31,
p. 703–714, Abril 2016.
FREDD, C. N.; DANIELS, J. L.; BAIHLY, J. D. $40 Billion Learning Curve:
Leveraging Lessons Learned to Minimize the Overall Investment in Unconventional
Plays. In: SPE Middle east unconventional resources conference and exhibition.
Society of Petroleum Engineers, 26 jan. 2015. <https://www.onepetro.org/conference-
paper/SPE-172973-MS>. Accessed on: 17 maio. 2016.
FUNDACIÓN JUBILEO. Situación de los Hidrocarburos en Bolivia: Serie Debate
Público No. 30. La Paz, Bolivia: Fundación Jubileo, 2014. .
<http://www.publishwhatyoupay.org/wp-
content/uploads/2015/04/Situacion_hidrocarburos_2014.pdf>. Accessed on: 8 set. 2015.
FURUOKA, F. Natural gas consumption and economic development in China and
Japan: An empirical examination of the Asian context. Renewable and Sustainable
Energy Reviews, v. 56, p. 100–115, Abril 2016.
FUSER, I. Conflitos e contratos: a petrobras, o nacionalismo boliviano e a
interdepedência do gás natural (2002-2010). 2011. Universidade de São Paulo, 2011.
171
<http://www.teses.usp.br/teses/disponiveis/8/8131/tde-03062013-125451/>. Accessed
on: 15 ago. 2016.
GABRIEL, S. A.; ROSENDAHL, K. E.; EGGING, R.; AVETISYAN, H. G.;
SIDDIQUI, S. Cartelization in gas markets: Studying the potential for a “Gas OPEC”.
Energy Economics, v. 34, n. 1, p. 137–152, jan. 2012.
GADANO, N. Determinantes de la inversión en el sector petróleo y gas en la
Argentina. Santiago, Chile: CEPAL, 1998. .
<http://repositorio.cepal.org/bitstream/handle/11362/7444/S9800088_es.pdf?sequence=
1>. Accessed on: 8 set. 2015.
GARRISON, J. Argentina Holds Vast Shale Reources. E&P Magazine, 4 jan. 2016.
<http://www.epmag.com/argentina-holds-vast-shale-resources-833866>. Accessed on:
17 jul. 2016.
GIL, S. Proyección de la demanda de gas para mediano y largo plazo. Petrotecnia, v.
48, n. 5, p. 86–101, 2007.
GIL, S. ¿Cómo se distribuye el consumo residencial de gas?. Modos de promover un
uso más eficiente. Petrotecnia, p. 86–92, 2013.
GLACHANT, J.-M.; HALLACK, M. Take-or-pay contract robustness: A three step
story told by the Brazil–Bolivia gas case? Energy Policy, v. 37, n. 2, p. 651–657, fev.
2009.
GNL QUINTERO. Reporte de Sustentabilidad 2014. Santiago, Chile: GNL Quintero,
2015. . <http://www.gnlquintero.com/pdf/reporte_sustentabilidad_GNLQ2014.pdf>.
Accessed on: 3 mar. 2016.
GONZÁLEZ, A. D.; CARLSSON-KANYAMA, A.; CRIVELLI, E. S.; GORTARI, S.
Residential energy use in one-family households with natural gas provision in a city of
the Patagonian Andean region. Energy Policy, v. 35, n. 4, p. 2141–2150, Abril 2007.
GRACCEVA, F.; ZENIEWSKI, P. Exploring the uncertainty around potential shale gas
development – A global energy system analysis based on TIAM (TIMES Integrated
Assessment Model). Energy, v. 57, p. 443–457, Agosto 2013.
HALLACK, M. Mecanismos de governo do comércio de gás entre brasil e bolívia.
2007. Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil, 2007.
<http://www.gee.ie.ufrj.br/index.php/get-tese-e-dissertacao/306-mecanismos-de-
governo-do-comercio-de-gas-entre-brasil-e-bolivia>. Accessed on: 15 ago. 2016.
HARDEN, C. (Jim). Discount Rate Development in Oil and Gas Valuation. In: SPE
hydrocarbon economics and evaluation symposium. Society of Petroleum Engineers,
19 maio 2014. <https://www.onepetro.org/conference-paper/SPE-169862-MS>.
Accessed on: 16 maio. 2016.
HARTLEY, P. R. The Future of Long-term LNG Contracts. The Energy Journal, v.
36, n. 3, 1 jul. 2015. <http://www.iaee.org/en/publications/ejarticle.aspx?id=2633>.
Accessed on: 10 ago. 2016.
172
HARTLEY, P. R.; KENNETH, M. B. The Baker Institute World Gas Trade Model.
Houston, USA: The James A. Baker III Institute for Public Policy, 2005. .
<http://bakerinstitute.org/media/files/Research/81966512/the-baker-institute-world-gas-
trade-model-biwgtm.pdf>. Accessed on: 28 mar. 2016.
HARTLEY, P. R.; MEDLOCK, III, K. B. Potential Futures for Russian Natural Gas
Exports. The Energy Journal, v. 30, n. 1, 1 set. 2009.
<http://www.iaee.org/en/publications/ejarticle.aspx?id=2317>. Accessed on: 10 ago.
2016.
HEAPS, C. G. Long-range energy alternatives planning (leap) system. Somerville,
MA, USA: Stockholm Environment Institute, 2012.
HINDERAKER, L.; NJAA, S. Utilization of Associated Petroleum Gas (APG) - The
Norwegian Experience (Russian). In: SPE Russian oil and gas conference and
exhibition. Society of Petroleum Engineers, 1 jan. 2010.
<https://www.onepetro.org/conference-paper/SPE-136316-RU>. Accessed on: 9 maio.
2016.
HOLZ, F.; RICHTER, P. M.; EGGING, R. A Global Perspective on the Future of
Natural Gas: Resources, Trade, and Climate Constraints. Review of Environmental
Economics and Policy, v. 9, n. 1, p. 85–106, 1 jan. 2015.
HOLZ, F.; VON HIRSCHHAUSEN, C.; KEMFERT, C. A strategic model of European
gas supply (GASMOD). Energy Economics, v. 30, n. 3, p. 766–788, Maio 2008.
HUBBERT, M. K. Nuclear energy and the fossil fuels. In: San Antonio, In: Meeting of
the southern district, division of production, AMERICAN PETROLEUM
INSTITUTE. San Antonio, Texas: 1956.
HUEBEL, R. R.; MALSAM, M. G. New NGL-recovery process provides viable
alternative. Oil and Gas Journal, 2012. <http://www.ogj.com/articles/print/vol-
110/issue-1a/processing/new-gl-recovery-p1.html>.
HUNTINGTON, H. G. Industrial natural gas consumption in the United States: An
empirical model for evaluating future trends. Energy Economics, Modeling of
Industrial Energy Consumption. v. 29, n. 4, p. 743–759, jul. 2007.
HUNTINGTON, H. G.; WEYANT, J. P.; SWEENEY, J. L. Modeling for insights, not
numbers: the experiences of the energy modeling forum. Omega, v. 10, n. 5, p. 449–
462, 1 jan. 1982.
HWANG, J.; LEE, K.-Y. Optimal liquefaction process cycle considering simplicity and
efficiency for LNG FPSO at FEED stage. Computers & Chemical Engineering, v. 63,
p. 1–33, Abril 2014.
HYDROCARBONS-TECHNOLOGY. Mejillones Liquefied Natural Gas (LNG)
Terminal, Chile. <http://www.hydrocarbons-technology.com/projects/cnlmejillones/>.
Accessed on: 14 jun. 2016.
IAPG. Producción de Petróleo - Datos Históricos.
<http://www.iapg.org.ar/estadisticasnew/historicospetroleopais.htm>.
173
IAPG. Informe anual 2015 de producción de petróleo y gas de reservorios no
convencionales de Neuquén. Petrotecnia, p. 26–44, 2016.
IBAMA. Ibama arquiva licenciamento da UHE São Luiz do Tapajós, no Pará.
<http://www.ibama.gov.br/publicadas/ibama-arquiva-licenciamento-da-uhe-sao-luiz-do-
tapajos-no-para>. Accessed on: 1 set. 2016.
IEA. World Energy Outlook 2011: Are we entering a golden age of natural gas?-
Special Report. Paris: OECD: International Energy Agency, 2011. .
<http://www.worldenergyoutlook.org/media/weowebsite/2011/WEO2011_GoldenAgeo
fGasReport.pdf>.
IEA. World Energy Investment Outlook. Paris, France: International Energy Agency,
2014a. .
<https://www.iea.org/publications/freepublications/publication/WEIO2014.pdf>.
IEA. Key World Energy Statistics 2014. Paris: OECD: International Energy Agency,
2014b. .
<http://www.iea.org/publications/freepublications/publication/KeyWorld2014.pdf>.
Accessed on: 4 nov. 2014.
IEA. Energy Technologies Perspectives 2015 - Mobilising Innovation to Accelerate
Climate Action. Paris, France: International Energy Agency, 2015a. .
IEA. Natural Gas Information: IEA Statistics. Paris, France: International Energy
Agency, 2015b. . . Accessed on: 12 out. 2015.
IEA. ETP-TIMES model for the energy conversion sector.
<https://www.iea.org/etp/etpmodel/energyconversion/>. Accessed on: 8 out. 2016a.
IEA. ETSAP-TIAM (TIMES Integrated Assessment Model). <http://www.iea-
etsap.org/web/applicationGlobal.asp>. Accessed on: 8 out. 2016b.
IEA-ETSAP. E-TechDS- Energy Technology Data Source: Oil and Gas Logistics.
<http://www.iea-etsap.org/Energy_Technologies/Energy_Supply.asp>. Accessed on: 20
maio. 2016.
IEA-ETSAP. Energy Technology System Analysis Program. <http://www.iea-
etsap.org/web/index.asp>. Accessed on: 20 maio. 2016.
IGU. World LNG Report - 2015 Edition: World Gas Conference Edition. Fornebu,
Norway: International Gas Union, 2015. . <http://www.igu.org/sites/default/files/node-
page-field_file/IGU-World%20LNG%20Report-2015%20Edition.pdf>. Accessed on:
26 dez. 2015.
IMF. World Economic Outlook Database.
<http://www.imf.org/external/pubs/ft/weo/2013/01/weodata/index.aspx>. Accessed on:
27 abr. 2015.
INDEC. Censo Nacional de Población, Hogares y Viviendas 2001.
<http://www.indec.gov.ar/micro_sitios/webcenso/censo2001s2/ampliada_index.asp?mo
de=01>. Accessed on: 5 fev. 2015.
174
INDEC. Censo Nacional de Población, Hogares y Viviendas 2010: Censo del
Bicentenario: Resultados definitivos, Serie B No.2 - 1aed. Buenos Aires: Instituto
Nacional de Estadística y Censos, 2012. .
INDEC. Proyecciones de población.
<http://www.indec.mecon.ar/nivel4_default.asp?id_tema_1=2&id_tema_2=24&id_tema
_3=84>. Accessed on: 5 fev. 2015.
INE. Anuario Estadístico 1999. La Paz, Bolivia: Instituto Nacional de Estadística,
1999. . <http://www.ine.gob.bo/pdf/publicaciones/ANUARIO99/Presentacion.pdf>.
INE. Censo 2002 Resultados: Volumen II Personas, Viviendas y Hogares. Santiago,
Chile: Instituto Nacional de Estadística, 2003. .
<http://www.ine.cl/canales/usuarios/cedoc_online/censos/pdf/censo_2002_volumen_II.
pdf>.
INE. Anuario Estadístico 2003. La Paz, Bolivia: Instituto Nacional de Estadística,
2004. . <http://www.ine.gob.bo/PDF/Anuario_2003/Anuario_2003.pdf>.
INE. Bolivia, Características de Población y Vivienda, Censo Nacional de
Poblacion y Vivienda 2012. La Paz, Bolivia: Instituto Nacional de Estadística, 2013a. .
<http://www.ine.gob.bo:8081/censo2012/PDF/resultadosCPV2012.pdf>.
INE. Síntesis de Resultados: Censo 2012. Santiago, Chile: Instituto Nacional de
Estadística, 2013b. . <http://www.iab.cl/wp-
content/themes/IAB/download.php?archivo=11803%7Cresumencenso_2012.pdf>.
INE. Anuarios Parque de Vehículos en Circulación.
<http://www.ine.cl/canales/chile_estadistico/estadisticas_economicas/transporte_y_com
unicaciones/parquevehiculos.php>. Accessed on: 8 out. 2015c.
INE. Estadísticas del Parque Automotor 2003-2013. La Paz, Bolivia: Instituto
Nacional de Estadística, 2014a. .
INE. Produccion de Petroleo y Gas natural.
<http://www.ine.gob.bo/indice/general.aspx?codigo=40105>.
IOGP. Environmental performance indicators - 2014 data. London, UK:
International Association of Oil & Gas Producers, 2015. .
<http://www.iogp.org/pubs/2014e.pdf>.
IPA. Perfiles de empresas productoras del sector petroquímico. Argentina: Instituto
Petroquímico Argentino, 2011. .
<http://www.ipa.org.ar/images/PUBLICACIONES/Perfilesempresasproductorassectorp
etroqumico.pdf>.
IRENA. Renewable Power Generation Costs in 2014. Bonn, Germany: International
Renewable Energy Agency, 2015. .
<http://www.irena.org/DocumentDownloads/Publications/IRENA_RE_Power_Costs_2
014_report.pdf>. Accessed on: 25 jan. 2016.
175
IRENA. Investment Opportunities in Latin America (Global Atlas).
<http://irena.masdar.ac.ae/?map=2012>. Accessed on: 19 ago. 2016.
JADRESIC, A. Investment in natural gas pipelines in the Southern Cone of Latin
America. [s.l.] The World Bank, 30 abr. 2000. .
<http://documents.worldbank.org/curated/en/117251468743744606/Investment-in-
natural-gas-pipelines-in-the-Southern-Cone-of-Latin-America>. Accessed on: 14 ago.
2016.
KANNAN, R. The development and application of a temporal MARKAL energy
system model using flexible time slicing. Applied Energy, v. 88, n. 6, p. 2261–2272,
jun. 2011.
KANORS. PET ( The Pan-European TIMES).
<http://www.kanors.com/Website/Mod_PET.asp>. Accessed on: 8 out. 2016.
KATHE, M. V.; EMPFIELD, A.; NA, J.; BLAIR, E.; FAN, L.-S. Hydrogen production
from natural gas using an iron-based chemical looping technology: Thermodynamic
simulations and process system analysis. Applied Energy, v. 165, p. 183–201, 1 mar.
2016.
KELMAN, R. Planejamento coordenado dos setores de energia elétrica e gás
natural. 2009. COPPE-UFRJ, Rio de Janeiro, Brazil, 2009. <http://www.psr-
inc.com/app/link/?t=d&f=tese_dsc_2009_rafael_kelman.pdf>. Accessed on: 15 ago.
2016.
KEMAUSUOR, F.; NYGAARD, I.; MACKENZIE, G. Prospects for bioenergy use in
Ghana using Long-range Energy Alternatives Planning model. Energy, v. 93, Part 1, p.
672–682, Dezembro 2015.
KEMP, A. G.; KASIM, S. The economics of CO2-EOR cluster developments in the UK
Central North Sea. Energy Policy, v. 62, p. 1344–1355, nov. 2013.
KEMP, J. New definitions needed for condensate and gas liquids. Reuters, 11 out.
2014. <http://www.reuters.com/article/us-condensate-usa-kemp-
idUSKCN0HZ03220141011>. Accessed on: 2 mar. 2016.
KHALILPOUR, R.; KARIMI, I. A. Selection of Liquefied Natural Gas (LNG)
Contracts for Minimizing Procurement Cost. Industrial & Engineering Chemistry
Research, v. 50, n. 17, p. 10298–10312, Setembro 2011.
KHORSHIDI, Z.; FLORIN, N. H.; HO, M. T.; WILEY, D. E. Techno-economic
evaluation of co-firing biomass gas with natural gas in existing NGCC plants with and
without CO2 capture. International Journal of Greenhouse Gas Control, v. 49, p.
343–363, jun. 2016.
KOZULJ, R. La industria del gas natural en América del Sur: situación y
posibilidades de la integración de mercados. Santiago, Chile: CEPAL, 2004. .
<http://www.cepal.org/es/publicaciones/6447-la-industria-del-gas-natural-en-america-
del-sur-situacion-y-posibilidades-de-la>. Accessed on: 14 ago. 2016.
176
KOZULJ, R. Situación y perspectivas del gas natural licuado en América del Sur.
Santiago, Chile: CEPAL, 2008. . <http://www.cepal.org/es/publicaciones/6330-
situacion-perspectivas-gas-natural-licuado-america-sur>. Accessed on: 14 ago. 2016.
KUMAR, S.; KWON, H.-T.; CHOI, K.-H.; LIM, W.; CHO, J. H.; TAK, K.; MOON, I.
LNG: An eco-friendly cryogenic fuel for sustainable development. Applied Energy, v.
88, n. 12, p. 4264–4273, Dezembro 2011.
LA NACION. Llegó el polémico buque regasificador.
<http://www.lanacion.com.ar/1017283-llego-el-polemico-buque-regasificador>.
Accessed on: 19 ago. 2016.
LA TERCERA. La opción por el gas vehicular. 19 ago. 2013.
<http://www.latercera.com/noticia/opinion/ideas-y-debates/2013/08/895-538372-9-la-
opcion-por-el-gas-vehicular.shtml>.
LA TERCERA. Gobierno informa que comenzó exportación de electricidad a
Argentina. 12 fev. 2016. <http://www.latercera.com/noticia/negocios/2016/02/655-
668124-9-gobierno-informa-que-comenzo-exportacion-de-electricidad-a-
argentina.shtml>. Accessed on: 3 mar. 2016.
LAHERRERE, J. Multi-Hubbert Modeling. jul. 1997. .
<http://www.oilcrisis.com/laherrere/multihub.htm>. Accessed on: 5 set. 2014.
LEE, D.-H.; HA, M.-K.; KIM, S.-Y.; SHIN, S.-C. Research of design challenges and
new technologies for floating LNG. International Journal of Naval Architecture and
Ocean Engineering, v. 6, n. 2, p. 307–322, jun. 2014.
LEMMERS, S. P. . LNG receiving and regasification terminal operations.
Hydrocarbon Processing, 2005.
<http://www.hydrocarbonprocessing.com/Article/2598903/LNG-receiving-and-
regasification-terminal-operations.html>.
LINDE. Natural Gas Processing Plants. Pullach, Germany: Linde Engineering, 2015.
. <http://www.linde-
engineering.com/internet.global.lindeengineering.global/en/images/HE_1_1_e_12_150d
pi19_4271.pdf>.
LION, M. B. Gas não convencional no brasil: condicionantes do desenvolvimento e
proposta de modelo de negócios. 2015. Universidade Federal do Rio de Janeiro, Rio
de Janeiro, Brazil, 2015. <http://www.ie.ufrj.br/images/pos-
graducao/pped/dissertacoes_e_teses/Manuella_Lion.pdf>. Accessed on: 15 ago. 2016.
LOCHNER, S.; BOTHE, D. From Russia with Gas. An analysis of the Nord Stream
pipeline´s impact on the European Gas Transmission System with the Tiger-
Model. Cologne, Germany: Institute of Energy Economics at the University of
Cologne, 2007. . <http://www.ewi.uni-
koeln.de/fileadmin/user_upload/Publikationen/Working_Paper/EWI_WP_07-02_Nord-
Stream-Impact.pdf>. Accessed on: 28 mar. 2016.
177
LOLOU, R.; GOLDSTEIN, G.; NOBLE, K. Documentation for the MARKAL
Family of Models. [s.l.] Energy Technology Systems Analysis Programme, 2004. .
<http://www.iea-etsap.org/web/mrkldoc-i_stdmarkal.pdf>.
LOLOU, R.; LEHTILA, A. Stochastic Programming and Tradeoff Analysis in
TIMES: Times Version 3.3 User Note. [s.l.] Energy Technology Systems Analysis
Programme, 2012. .
LOLOU, R.; REMNE, U.; KANUDIA, A.; LEHTILA, A.; GOLDSTEIN, G.
Documentation for the TIMES Model - PART I. [s.l.] Energy Technology Systems
Analysis Programme, 2005. . <http://www.iea-etsap.org/web/docs/timesdoc-intro.pdf>.
LUCCHESI, C. F. Petróleo. Estudos Avançados, v. 12, n. 33, p. 17–40, ago. 1998.
LUCENA, A. F. P.; SZKLO, A. S.; SCHAEFFER, R.; DE SOUZA, R. R.; BORBA, B.
S. M. C.; DA COSTA, I. V. L.; JÚNIOR, A. O. P.; DA CUNHA, S. H. F. The
vulnerability of renewable energy to climate change in Brazil. Energy Policy, v. 37, n.
3, p. 879–889, mar. 2009.
MAGGIO, G.; CACCIOLA, G. A variant of the Hubbert curve for world oil production
forecasts. Energy Policy, v. 37, n. 11, p. 4761–4770, nov. 2009.
MARES, D. R. Natural gas pipelines in the Southern Cone. In: Natural Gas and
Geopolitics. [s.l.] Cambridge University Press, 2006.
MARES, D. R.; MARTIN, J. M. Regional Energy Integration in Latin America: lessons
from Chile’s experience with natural gas. Third World Quarterly, v. 33, n. 1, p. 55–
70, Fevereiro 2012.
MARTINS, M. R.; PESTANA, M. A.; SOUZA, G. F. M.; SCHLEDER, A. M.
Quantitative risk analysis of loading and offloading liquefied natural gas (LNG) on a
floating storage and regasification unit (FSRU). Journal of Loss Prevention in the
Process Industries, 2016.
<http://www.sciencedirect.com/science/article/pii/S095042301630208X>. Accessed on:
19 ago. 2016.
MATHIAS, M. C.; SZKLO, A. Lessons learned from Brazilian natural gas industry
reform. Energy Policy, v. 35, n. 12, p. 6478–6490, Dezembro 2007.
MAZYAN, W.; AHMADI, A.; AHMED, H.; HOORFAR, M. Market and technology
assessment of natural gas processing: A review. Journal of Natural Gas Science and
Engineering, v. 30, p. 487–514, mar. 2016.
MCGLADE, C. E. Uncertainties in the outlook for oil and gas. 2013. University
College London, London, UK, 2013.
<http://discovery.ucl.ac.uk/1418473/2/131106%20Christophe%20McGlade_PhD%20T
hesis.pdf>. Accessed on: 25 jan. 2016.
MCGLADE, C.; EKINS, P. The Geographical Distribution of Fossil Fuels Unused
When Limiting Global Warming to 2 °C. Nature, v. 517, n. 7533, p. 187–190, 8 jan.
2015.
178
MCGLADE, C.; WATSON, J.; BRADSHAW, M.; PYE, S.; EKINS, P. The future
role of natural gas in the UK. London, UK: UKERC, 2016. .
<http://www.ukerc.ac.uk/publications/the-future-role-of-natural-gas-in-the-uk.html>.
Accessed on: 14 ago. 2016.
MEDLOCK III, K. B.; JAFFE, A. M.; HARTLEY, P. R. Shale Gas and U.S. National
Security. Houston, USA: The James A. Baker III Institute for Public Policy, 2011. .
<http://bakerinstitute.org/media/files/Research/ccecf6b6/EF-pub-DOEShaleGas-
07192011.pdf>. Accessed on: 28 mar. 2016.
MHE. Plan Óptimo de Expansión del Sistema Interconectado Nacional 2012-2022.
La Paz, Bolivia: Ministerio de Hidrocarburos y Energía del estado Plurinacional de
Bolivia, 2012. .
<http://www2.hidrocarburos.gob.bo/phocadownload/Plan%20de%20Expansi%C3%B3n
%20del%20SIN%202012-2022.pdf>.
MHE. Balance Energético Nacional. La Paz, Bolivia: Ministerio de Hidrocarburos y
Energía del estado Plurinacional de Bolivia, 2013. .
MHE. Balance Energético Nacional 2000-2014. La Paz, Bolivia: Ministerio de
Hidrocarburos y Energía del estado Plurinacional de Bolivia, 2015. .
MIEM. Acuerdo Energético entre la República Oriental del Uruguay y la
República Argentina. Montevideo, Uruguay: Ministerio de Industria, Energía y
Minería, 9 maio 2016. .
<https://medios.presidencia.gub.uy/legal/2016/resoluciones/05/miem_90.pdf>.
MIEMDNETN. Estudios de base para el diseño de estrategias y políticas
energéticas: relevamiento de consumos de energía sectoriales en términos de
energía útil a nivel nacional. Montevideo, Uruguay: Dirección Nacional de Energía y
Tecnología Nuclear, 2009. .
<http://www.dne.gub.uy/documents/15377/40852/Informe%20Sector%20Industrial.pdf
>. Accessed on: 6 nov. 2015.
MINEM. Precios de Comercio Exterior.
<http://www.energia.gob.ar/contenidos/verpagina.php?idpagina=3289>. Accessed on:
22 maio. 2015.
MINENERGIA. Balance Nacional de Energía 2012. Santiago de Chile: Ministerio de
Energía de Chile, 2013. .
<http://antiguo.minenergia.cl/minwww/opencms/14_portal_informacion/06_Estadistica
s/Balances_Energ.html>. Accessed on: 9 nov. 2015.
MINENERGIA. Hoja de Ruta 2050: Hacia una energía sustentable e inclusiva para
Chile. Comité Consultivo de Energía 2050. Santiago de Chile: Ministerio de Energía
de Chile, 2015. . <http://www.energia2050.cl/uploads/libros/hojaderuta.pdf>. Accessed
on: 9 nov. 2015.
MME. Boletim de Exploração e Produção de Petróleo e Gás Natural, Edição 47.
Brasilia: Ministério de Minas e Energia, 2016a. .
<http://www.mme.gov.br/documents/10584/2181192/Boletim+de+Explora%C3%A7%
C3%A3o+e+Produ%C3%A7%C3%A3o+de+Petr%C3%B3leo+e+G%C3%A1s+Natura
179
l+n%C2%BA+47+-+mar%C3%A7o/6660019d-fa47-48c8-90d8-
39728d4c2edf?version=1.0>. Accessed on: 5 abr. 2016.
MME. Boletim Mensal de Acompanhamento da Indústria de Gás Natural, Edição
No. 107. Brasilia: Ministério de Minas e Energia, 2016b. .
<http://www.mme.gov.br/documents/1138769/1732803/Boletim_Gas_Natural_nr_107_
JAN_16.pdf/9a2fcf86-b7fe-445e-a481-6a58b7e1b689>. Accessed on: 5 abr. 2016.
MOURA, G.; HOWELLS, M. SAMBA, The open source South American Model
Base. A Brazilian Perspective on Long Term Power Systems Investment and
Integration: Work. Pap. Ser., no. DESA /15/8/11. Stockholm: Royal Institute of
Technology (KTH), 2015. . <http://www.osemosys.org/samba-south-american-model-
base.html>.
MUÑOZ, C.; INOSTROZA, J. R.; GALETOVIC, A. Gas y electricidad: ¿Qué hacer
ahora? Revista de Estudios Públicos, v. 96, 2004.
<http://www.cepchile.cl/cep/site/artic/20160304/pags/20160304093445.html>.
Accessed on: 18 ago. 2016.
NIJBOER, M. The contribution of natural gas vehicles to sustainable transport.
Paris, France: International Energy Agency, 2010. .
NUESTROMAR. Regasificación polémica en Bahía Blanca.
<http://www.nuestromar.org/noticias/puertos_04_2008_regasificacion_polemica_en_ba
hia_blanca_15714>. Accessed on: 19 ago. 2016.
OGJ. Brazil’s ANP, Petrobras lock horns over gas flaring in Campos basin. Oil and
Gas Journal, 29 jul. 2002. <http://www.ogj.com/articles/print/volume-100/issue-
30/general-interest/brazils-anp-petrobras-lock-horns-over-gas-flaring-in-campos-
basin.html>.
OGJ. Capital expenditures to be squeezed further in 2016.
<http://www.ogj.com/articles/print/volume-114/issue-3/special-report-capital-spending-
outlook/capital-expenditures-to-be-squeezed-further-in-2016.html>. Accessed on: 1 jul.
2016.
ONS. Histórico da Operação: Geração de Energia.
<http://www.ons.org.br/historico/geracao_energia.aspx>. Accessed on: 27 abr. 2016.
ONS; SORIA, R. Curva Carga Global, dados entregados pela ONS, 2014. .
OWEN, N. A.; INDERWILDI, O. R.; KING, D. A. The status of conventional world oil
reserves—Hype or cause for concern? Energy Policy, v. 38, n. 8, p. 4743–4749, Agosto
2010.
PALTSEV, S. Scenarios for Russia’s natural gas exports to 2050. Energy Economics,
v. 42, p. 262–270, mar. 2014.
PDVSA. Presentadas líneas del Plan Estratégico Socialista de PDVSA 2016-2025.
<http://www.pdvsa.com/index.php?tpl=interface.sp/design/salaprensa/readnew.tpl.html
&newsid_obj_id=13356&newsid_temas=1>. Accessed on: 28 abr. 2016.
180
PETROBRAS. Unidade de Processamento de Gás Natural Boliviano e
investimentos no Estado do MS2013, Brasilia, Brazil. .
<http://www19.senado.gov.br/sdleg-getter/public/getDocument?docverid=499d4385-
f9c6-4070-b042-7c23bccb532c;1.0>. Accessed on: 3 abr. 2016.
PETROBRAS. Província Petrolífera de Urucu: 20 anos.
<http://sites.petrobras.com.br/minisite/urucu/urucu.html>. Accessed on: 20 jan. 2016.
PETROBRAS. Plano de Negócio e Gestão: 2013-2017. Rio de Janeiro, Brazil:
Petróleo Brasileiro S.A., 2013. .
<http://bibspi.planejamento.gov.br/bitstream/handle/iditem/222/PETROBRAS_PNG_2
013_2017.pdf?sequence=1>. Accessed on: 20 jan. 2016.
PETROBRAS. 2030 Strategic Plan and 2014-2018 Business and Management Plan -
Presentation. [s.l: s.n.]. <http://www.investidorpetrobras.com.br/en/business-
management-plan/2030-strategic-plan-and-2014-2018-business-and-management-plan-
presentation.htm>.
PETROBRAS. Iniciamos Operação do Terceiro Terminal de Regaseificação de
GNL. <http://fatosedados.blogspetrobras.com.br/2014/01/24/iniciamos-operacao-do-
terceiro-terminal-de-regaseificacao-de-gnl/>. Accessed on: 20 jan. 2016b.
PETROBRAS. Urucu-Coari-Manaus. <http://www.petrobras.com.br/pt/nossas-
atividades/principais-operacoes/gasodutos/urucu-coari-manaus.htm>. Accessed on: 20
jan. 2016a.
PETROBRAS. Plano de Negócio e Gestão: 2015-2019. Rio de Janeiro, Brazil:
Petróleo Brasileiro S.A., 2015b. .
<http://www.investidorpetrobras.com.br/download/3013>. Accessed on: 20 jan. 2016.
PETROBRAS. Annual Report on FORM 20-F 2015. Rio de Janeiro, Brazil: Petróleo
Brasileiro S.A., 2016. .
<https://www.sec.gov/Archives/edgar/data/1119639/000129281416004364/pbraform20
f_2015.htm>. Accessed on: 20 jun. 2016.
PFENNINGER, S.; HAWKES, A.; KEIRSTEAD, J. Energy systems modeling for
twenty-first century energy challenges. Renewable and Sustainable Energy Reviews,
v. 33, p. 74–86, Maio 2014.
PIZARRO, J. O. D. S.; BRANCO, C. C. M. Challenges in Implementing an EOR
Project in the Pre-Salt Province in Deep Offshore Brasil. In: SPE EOR conference at
oil and gas West Asia. Society of Petroleum Engineers, 1 jan. 2012.
<https://www.onepetro.org/conference-paper/SPE-155665-
MS?sort=&start=10&q=associated+gas+pre-
salt&from_year=&peer_reviewed=&published_between=&fromSearchResults=true&to
_year=&rows=10#>. Accessed on: 9 maio. 2016.
PLANALTO. DECRETO No 3.371, DE 24 DE FEVEREIRO DE 2000.
<http://www.planalto.gov.br/ccivil_03/decreto/D3371.htm>.
181
PLATTS. Oilgram price report. Usa: PLATTS Mc Graw Hill Financial, 29 set. 2015. .
<https://www.platts.com/IM.Platts.content/ProductsServices/Products/oilgrampricerepo
rt.pdf>.
PORTUGAL-PEREIRA, J.; KÖBERLE, A. C.; SORIA, R.; LUCENA, A. F. P.;
SZKLO, A.; SCHAEFFER, R. Overlooked impacts of electricity expansion
optimisation modelling: The life cycle side of the story. Energy, [s.d.]
<http://www.sciencedirect.com/science/article/pii/S0360544216303061>. Accessed on:
11 maio. 2016.
POSTIC, S. Long-term energy prospective modeling for south america -
applications to international climate negotiations. 2015. Ecole Nationale Supérieure
des Mines de Paris, Paris, France, 2015. <https://pastel.archives-ouvertes.fr/tel-
01303955/document>. Accessed on: 20 maio. 2016.
PULIAFITO, S. E.; ALLENDE, D.; PINTO, S.; CASTESANA, P. High resolution
inventory of GHG emissions of the road transport sector in Argentina. Atmospheric
Environment, v. 101, p. 303–311, Enero 2015.
PWC. Economic projections. <http://www.pwc.com/gx/en/issues/economy/global-
economy-watch/projections.jhtml>. Accessed on: 27 abr. 2015.
RAFINDADI, A. A.; OZTURK, I. Natural gas consumption and economic growth
nexus: Is the 10th Malaysian plan attainable within the limits of its resource?
Renewable and Sustainable Energy Reviews, v. 49, p. 1221–1232, Setembro 2015.
RBN ENERGY. Fifty shades of condensates. Oil and Gas Financial Journal, 2012.
<http://www.ogfj.com/articles/2012/10/fifty-shades.html>. Accessed on: 2 mar. 2016.
REACCESS. REACCESS - Risk of Energy Availability Common Corridors for
Europe Supply Security. [s.l: s.n.]. <http://reaccess.epu.ntua.gr/>.
REMME, U.; GOLDSTEIN, G. A.; SCHELLMANN, U.; SCHLENZIG, C.
MESAP/TIMES - advanced decision support for energy and environmental planning.
In: Operations Research Proceedings 2001, Duisburg, Germany. In: International
conference on operations research (OR 2001). Duisburg, Germany: Springer, 2001.
REVISTAPETROQUIMICA. YPF piensa en alternativas para agregar valor a Vaca
Muerta, 10 set. 2015. . <http://revistapetroquimica.com/ypf-piensa-en-alternativas-
para-agregar-valor-a-vaca-muerta/>. Accessed on: 27 out. 2015.
REYNOLDS, D. B.; BAEK, J. Much ado about Hotelling: Beware the ides of Hubbert.
Energy Economics, v. 34, n. 1, p. 162–170, jan. 2012.
RÍOS-MERCADO, R. Z.; BORRAZ-SÁNCHEZ, C. Optimization problems in natural
gas transportation systems: A state-of-the-art review. Applied Energy, v. 147, p. 536–
555, 1 jun. 2015.
RISUELO, F. Análisis de la infraestructura de gas natural en la República
Argentina. Buenos Aires, Argentina: Cámara Argentina de la Construcción, 2010. .
<http://www.camarco.org.ar/File/GetPublicFile?id=516>. Accessed on: 8 set. 2015.
182
RITZ, R. A. Price discrimination and limits to arbitrage: An analysis of global LNG
markets. Energy Economics, v. 45, p. 324–332, Setembro 2014.
ROCHEDO, P. R. R.; COSTA, I. V. L.; IMPÉRIO, M.; HOFFMANN, B. S.;
MERSCHMANN, P. R. de C.; OLIVEIRA, C. C. N.; SZKLO, A.; SCHAEFFER, R.
Carbon capture potential and costs in Brazil. Journal of Cleaner Production, v. 131, p.
280–295, Setembro 2016.
RODRIGUES, G. M. S.; OLIVEIRA, G. A. Overview of the Natural Gas Industry in
the Southern Cone. In: 26TH World gas conference, 1-5 JUNE 2015. Paris, France:
International Gas Union, 2015.
RODRIGUES, N.; COLOMER, M. The crisis in the oil industry and its impacts on the
Brazilian macroeconomic indicators. In: 39TH IAEE International conference.
Bergen, Norway: International Association of Energy Economics, 2016.
<http://www.iaee.org/iaee2016/submissions/ExtendedAbs/Ni%C3%A1gara%20Rodrig
ues.pdf>. Accessed on: 9 jul. 2016.
RODRÍGUEZ, J. J. Operaciones de Regasificación de GNL en Argentina. In:
JORNADAS DE PRODUCCIÓN, TRANSPORTE Y TRATAMIENTO DE GAS: EL
DESAFÍO DEL GAS NO CONVENCIONAL. Buenos Aires, Argentina., 31 ago. 2011.
.
<http://www.iapg.org.ar/sectores/eventos/eventos/listados/presentacionesjornadas/21.pd
f>. Accessed on: 11 maio. 2015.
ROJAS, L. El futuro del petróleo y del gas natural en Chile. In: TERCER
SEMINARIO MES DE LA ENERGÍA: “PETRÓLEO Y GAS”. Santiago, Chile, 2012. .
<http://www.ingenieros.cl/tercer-seminario-petroleo-y-gas/>.
RRC. Railroad Comission of Texas: RULE §3.79. . 2003.
RUDNICK, H.; BARROSSO, L.; CUNHA, G.; MOCARQUER, S. A Natural Fit:
Electricity-Gas Integration Challenges in South America. Power and Energy
Magazine, IEEE, v. 12, n. 6, p. 29–39, 2014.
SÄLLH, D.; WACHTMEISTER, H.; TANG, X.; HÖÖK, M. Offshore oil: Investigating
production parameters of fields of varying size, location and water depth. Fuel, v. 139,
p. 430–440, 1 jan. 2015.
SAMUELSON, P. A. Spatial Price Equilibrium and Linear Programming. The
American Economic Review, v. 42, n. 3, p. 283–303, 1952.
SÁNCHEZ-ÚBEDA, E. F.; BERZOSA, A. Modeling and forecasting industrial end-use
natural gas consumption. Energy Economics, Modeling of Industrial Energy
Consumption. v. 29, n. 4, p. 710–742, jul. 2007.
SANTOS, R. M. Alternativas de monetização de recursos de gás natural em terra:
o caso da bacia do paraná. 2015. Universidade Federal do Rio de Janeiro, Rio de
Janeiro, Brazil, 2015.
<http://www.ppe.ufrj.br/ppe/production/tesis/ricardo_moreira.pdf>.
183
SANTOS, E. M. dos; MAZAFERRO, M. A. N.; OXILIA, V. Trans-boundary Industrial
Integration between Bolivia and Brazil. Oil, Gas & Energy Law Journal (OGEL), v.
2, n. 4, 1 out. 2004. <https://www.ogel.org/article.asp?key=1562>. Accessed on: 15
ago. 2016.
SARAIVA, T. A.; SZKLO, A.; LUCENA, A. F. P.; CHAVEZ-RODRIGUEZ, M. F.
Forecasting Brazil’s crude oil production using a multi-Hubbert model variant. Fuel, v.
115, p. 24–31, Enero 2014.
SBM OFFSHORE. Offshore application of Gas to Liquid Technology to cut flaring
on FPSOs. In: GASETECH, CONFERENCE & EXHIBITION. Seoul, Korea, 2014. .
<http://www.compactgtl.com/wp-content/uploads/2015/01/Presentation-Joint-with-
SBM-Gastech-Seoul-March-2014.pdf>. Accessed on: 3 maio. 2016.
SCHMIDT, J.; CANCELLA, R.; JUNIOR, A. O. P. The effect of windpower on long-
term variability of combined hydro-wind resources: The case of Brazil. Renewable and
Sustainable Energy Reviews, v. 55, p. 131–141, mar. 2016.
SCHNEIDER, L. El Proyecto de Regasificación de GNL en Argentina. Buenos
Aires, Argentina: Centro Latinoamericano de Investigaciones Científicas y Técnicas,
2008. . <http://www.cienciayenergia.com/Contenido/pdf/080601_ls_arg.pdf>.
Accessed on: 8 set. 2015.
SEC. Informe Estadístico.
<http://www.sec.cl/portal/page?_pageid=33,3429539&_dad=portal&_schema=PORTA
L>.
SECRETARIA DE ENERGIA. Resolución 265/2004. . 2004.
SECRETARIA DE ENERGIA. Informe Estadístico del Sector Eléctrico 2012.
<http://www.energia.gov.ar/contenidos/verpagina.php?idpagina=3783>. Accessed on:
10 abr. 2015.
SECRETARIA DE ENERGIA. Balance Energético Nacional.
<http://www.energia.gov.ar/contenidos/verpagina.php?idpagina=3366>. Accessed on:
10 abr. 2015a.
SECRETARIA DE ENERGIA. Producción de Petróleo y Gas (Tablas Dinámicas).
<http://energia3.mecon.gov.ar/contenidos/verpagina.php?idpagina=3299>. Accessed
on: 24 dez. 2015b.
SECRETARIA DE ENERGIA. Datos de Reservas Comprobadas y Probables de
Petróleo y Gas hasta el Final de la Vida Útil de los Yacimientos.
<http://energia3.mecon.gov.ar/contenidos/verpagina.php?idpagina=3312>. Accessed
on: 24 dez. 2015c.
SECRETARIA DE ENERGIA. Plan de Acción e Inversiones a Ejecutar (Tablas
Dinámicas).
<http://energia3.mecon.gov.ar/contenidos/verpagina.php?idpagina=3331>. Accessed
on: 24 dez. 2015d.
184
SEDDON, D. Gas usage & value: the technology and economics of natural gas use
in the process industries. Tulsa, Oklahoma: Pennwell Publishing Company, 2006.
SENA, M. F. M. de; ROSA, L. P.; SZKLO, A. Will Venezuelan extra-heavy oil be a
significant source of petroleum in the next decades? Energy Policy, v. 61, p. 51–59,
out. 2013.
SGI. Methane and CO2 Emissions from the Natural Gas Supply Chain: An
Evidence Assessment. London, UK: Sustainable Gas Institute, 2015. .
<http://www.sustainablegasinstitute.org/wp-
content/uploads/2015/09/SGI_White_Paper_methane-and-CO2-emissions_WEB-
FINAL.pdf?noredirect=1>. Accessed on: 4 mar. 2016.
SGI. Sustainable Gas Institute Upstream database. London, UK: Imperial College
London, 2016. .
SILVA, M. I. M. T. da. Estudo de mitigação de gases de efeito estufa na produção
de petróleo brasileira - ênfase offshore. 2015. UFRJ/COPPE Programa de
Planejamento Energético, Rio de Janeiro, 2015.
<http://www.ppe.ufrj.br/ppe/production/tesis/imperio.pdf>. Accessed on: 25 jan. 2016.
SIMOES, S.; NIJS, W.; CASTELLO, P.; SGOBBI, A.; RADU, D.; BOLAT, P.;
THIEL, C.; PETEVES, E. The JRC-EU-TIMES model - Assessing the long-term
role of the SET Plan Energy technologies: Institute for Energy and Transport. [s.l.]
JRC Institutes, 2013. .
<http://publications.jrc.ec.europa.eu/repository/handle/JRC85804>.
SIMS, R.; SCHAEFFER, R.; CREUTZIG, X.; CRUZ-NÚÑEZ, M.; D´AGOSTO, D.;
DIMITRIU, M. J.; MEZA, L. F.; FULTON, S.; KOBAYASHI, O.; LAH, A.;
MCKINNON, P.; NEWMAN, M.; OUYANG, J.; SCHAUER, D.; SPERLING;
TIWARI, G. Transport. Climate Change 2014: Mitigation of Climate Change.
Contribution of Working Group III to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. New York, USA: International Panel
of Climate Change, 2014. . <https://www.ipcc.ch/pdf/assessment-report/ar4/wg3/ar4-
wg3-chapter4.pdf>.
SMITH, J. L. Valuing Barrels of Oil Equivalent. The Energy Journal, v. 36, n. 1, 1 set.
2015. <http://www.iaee.org/en/publications/ejarticle.aspx?id=2668>. Accessed on: 27
fev. 2016.
SMITH, T. R.; SU, J.; EL ABBADI, A.; AGRAWAL, D.; ALONSO, G.; SARAN, A.
Computational modeling systems. Information Systems, Scientific Databases. v. 20, n.
2, p. 127–153, Abril 1995.
SOLDO, B. Forecasting natural gas consumption. Applied Energy, v. 92, p. 26–37,
Abril 2012.
SORIA, R. REMix-CEM_BRAZIL: Overview Status of Work. Rio de Janeiro:
COPPE/DLR, 2016. .
SOUSA-AGUIAR, E. F.; APPEL, L. G.; MOTA, C. Natural gas chemical
transformations: The path to refining in the future. Catalysis Today, Gas-to-Liquid
185
TechnologyPapers presented at the 12th Brazilian Congress on Catalysis. v. 101, n. 1, p.
3–7, 15 mar. 2005.
SPE. Petroleum Resources Management System. [s.l.] Society of Petroleum
Engineers, 2007. .
<http://www.spe.org/industry/docs/Petroleum_Resources_Management_System_2007.p
df>. Accessed on: 27 ago. 2015.
SPE. Guidelines for Application of the Petroleum Resources Management System.
[s.l.] Society of Petroleum Engineers, 2011. .
<http://www.spe.org/industry/docs/PRMS_Guidelines_Nov2011.pdf>. Accessed on: 27
ago. 2015.
SRR. Understanding SEC Oil and Gas Reserve Reporting. [s.l.] Stout Risius Ross,
2015. .
<http://www.iea.org/publications/freepublications/publication/ngl2010_free.pdf>.
STEFANA, E.; MARCIANO, F.; ALBERTI, M. Qualitative risk assessment of a Dual
Fuel (LNG-Diesel) system for heavy-duty trucks. Journal of Loss Prevention in the
Process Industries, v. 39, p. 39–58, jan. 2016.
SUGANTHI, L.; SAMUEL, A. A. Energy models for demand forecasting—A review.
Renewable and Sustainable Energy Reviews, v. 16, n. 2, p. 1223–1240, Fevereiro
2012.
SUTHERLAND, R. J. Natural gas contracts in an emerging competitive market.
Energy Policy, v. 21, n. 12, p. 1191–1204, Dezembro 1993.
SZKLO, A.; MACHADO, G.; SCHAEFFER, R. Future oil production in Brazil—
Estimates based on a Hubbert model. Energy Policy, v. 35, n. 4, p. 2360–2367, Abril
2007.
SZKLO, A. S.; CARNEIRO, J. T. G.; MACHADO, G. Break-even price for upstream
activities in Brazil: Evaluation of the opportunity cost of oil production delay in a non-
mature sedimentary production region. Energy, v. 33, n. 4, p. 589–600, Abril 2008.
TGS. Capacidades y Datos: El complejo de General Cerri.
<http://www.tgs.com.ar/Servicios/Procesamiento/Capacidades-y-datos>. Accessed on:
14 jun. 2016.
THOMAS, J. E. Fundamentos de engenharia de petroleo. 2da. ed. [s.l.] Interciência,
2004.
UN-DESA. World Population Prospects: The 2012 Revision.
<http://esa.un.org/unpd/wpp/Excel-Data/population.htm>. Accessed on: 23 fev. 2015.
UPME. Plan de Expansión de Referencia Generación – Transmisión 2012-2025.
[s.l.] Unidad de Planeación Minero Energética, 2012. .
<http://www.siel.gov.co/LinkClick.aspx?fileticket=ZzUcD1SYL7U%3D&tabid=57>.
186
USGS. An Estimate of Undiscovered Conventional Oil and Gas Resources of the
World, 2012: World Petroleum Resources Project. [s.l.] U.S. Geological Survey, 2012.
. <http://pubs.usgs.gov/fs/2012/3042/fs2012-3042.pdf>.
USGS. Supporting Data for the U.S. Geological Survey 2012 World Assessment of
Undiscovered Oil and Gas Resources. <http://pubs.usgs.gov/dds/dds-069/dds-069-
ff/>.
USGS. Assessment of Unconventional Tight-Gas Resources of the Magallanes Bain
Province, Chile 2015: U.S. Geological Survery Energy Resources Program.
Washington, DC, USA: U.S. Geological Survey, 2016. .
<http://pubs.usgs.gov/fs/2015/3085/fs20153085.pdf>. Accessed on: 3 mar. 2016.
UWITONZE, H.; LEE, I.; HWANG, K. S. Alternatives of integrated processes for
coproduction of LNG and NGLs recovery. Chemical Engineering and Processing:
Process Intensification, [s.d.]
<http://www.sciencedirect.com/science/article/pii/S0255270116301222>. Accessed on:
14 jun. 2016.
VAGGE, C.; CZAJKOWSKI, J.; CELINA, F. Análisis del consumo de gas natural en
una vivienda en la ciudad de La Plata. Avances en Energías Renovables y Medio
Ambiente, v. 12, p. 05.81-05.87, 2008.
VAN DEN BROEK, M.; BERGHOUT, N.; RUBIN, E. S. The potential of renewables
versus natural gas with CO2 capture and storage for power generation under CO2
constraints. Renewable and Sustainable Energy Reviews, v. 49, p. 1296–1322,
Setembro 2015.
VARGAS SALGUEIRO, A. Y.p.f.b. entre nacionalistas y liberales. La Paz, Bolivia:
Los amigos del Libro, 1996.
VÁSQUEZ, P. Argentina´s Oil and Gas Sector. Washington, DC, USA: Wilson
Center, 2016. .
<https://www.wilsoncenter.org/sites/default/files/argentinasoilgas.vasquez.pdf>.
Accessed on: 29 jun. 2016.
VERMA, M. K. Fundamentals of Carbon Dioxide-Enhanced Oil Recovery (CO2-
EOR) - A Supporting Document of the Assessment Methodology for Hydrocarbon
Recovery Using CO2-EOR Associated with Carbon Sequestration. Reston,
Virginia: U.S. Geological Survey, 2015. .
<http://pubs.usgs.gov/of/2015/1071/pdf/ofr2015-1071.pdf>.
VIEGAS, T. de O. C. Competitividade em custos na atividade petrolífera em águas
profundas. 2013. UFRJ/Instituto de Economía, Rio de Janeiro, 2013.
<http://www.gee.ie.ufrj.br/index.php/get-tese-e-dissertacao/539-competitividade-em-
custos-na-atividade-petrolifera-em-aguas-profundas>. Accessed on: 25 jan. 2016.
WANG, T.; LIN, B. China’s natural gas consumption and subsidies—From a sector
perspective. Energy Policy, v. 65, p. 541–551, Fevereiro 2014.
WEIMER-JEHLE, W.; BUCHGEISTER, J.; HAUSER, W.; KOSOW, H.; NAEGLER,
T.; POGANIETZ, W.-R.; PREGGER, T.; PREHOFER, S.; VON
187
RECKLINGHAUSEN, A.; SCHIPPL, J.; VÖGELE, S. Context scenarios and their
usage for the construction of socio-technical energy scenarios. Energy, v. 111, p. 956–
970, Setembro 2016.
WOOD, D. A. Competition is driving significant change in natural gas LNG and
pipeline export markets. Journal of Natural Gas Science and Engineering, v. 33, p.
A1–A5, jul. 2016.
WORLD BANK. Industry, value added (% of GDP).
<http://data.worldbank.org/indicator/NV.IND.TOTL.ZS?page=6>.
YPF. Plan Estratégico. <https://www.ypf.com/EnergiaYPF/Paginas/plan-
estrategico.html>. Accessed on: 17 maio. 2016.
YPF. FORM 20-F, ANNUAL REPORT PURSUANT TO SECTION 13 OR 15, OF
THE SECURITIES EXCHANGE ACT OF 1934,For the fiscal year ended
December 31, 2014. Buenos Aires, Argentina: YPF Sociedad Anónima, 2015. .
<http://www.sec.gov/Archives/edgar/data/904851/000119312515111986/d863353d20f.
htm>. Accessed on: 6 nov. 2015.
YPF. FORM 20-F, ANNUAL REPORT PURSUANT TO SECTION 13 OR 15, OF
THE SECURITIES EXCHANGE ACT OF 1934,For the fiscal year ended
December 31, 2015. Buenos Aires, Argentina: YPF Sociedad Anónima, 2016. .
<https://www.ypf.com/enu/InversoresAccionistas/Lists/InformeAnualForm20/YPF-
20F-2015.PDF>. Accessed on: 6 nov. 2015.
YPFB. Reservas de gas y petróleo suben y garantizan mercado interno y exportación a
2025. CAMBIO, p. 24, 23 jul. 2014.
YPFB. YPFB programó $US 7.459 millones de inversión en exploración y
explotación. <http://www.ypfb.gob.bo/en/14-noticias/379-ypfb-program%C3%B3-
$us-7-459-millones-de-inversi%C3%B3n-en-exploraci%C3%B3n-y-
explotaci%C3%B3n.html>. Accessed on: 17 maio. 2016a.
YPFB. Boletín Estadístico Gestión 2014. La Paz, Bolivia: Yacimientos Petrolíferos
Fiscales Bolivianos, 2015b. .
<http://www.ypfb.gob.bo/es/component/phocadownload/category/86-
boletines.html?download=955:bolet%C3%ADn-estad%C3%ADstico-2014>.
YPFB. Boletínes Estadísticos Gestiones 2009 a 2014.
<http://www.ypfb.gob.bo/es/55-ypfb-corporacion/publicaciones/79-
publicaciones.html>.
YPFB. Presentación YPFB Corporación en Audiencia Pública de Rendición de
Cuentas final 2014 e inicial 2015La Paz, Bolivia, 17 jan. 2015d. .
<http://www2.hidrocarburos.gob.bo/index.php/transparencia/audiencia-
p%C3%BAblica/category/49-audiencia-publica-de-rendicion-de-cuentas-final-2014-e-
inicial-2015.html>.
188
10 ANNEX
10.1 ANNEX A: NATURAL GAS DEMAND SUPPLEMENTARY MATERIAL
Annex A presents the supplementary material regarding related to some modelling
issues of the natural gas demand of end-uses.
Natural gas demand split in North Chile and Central-South Chile
Chile was divided in two regions: North Chile and Central-South Chile. North Chile
involves Arica and Parinacota regions down to Antofagasta; Central-South Chile involves from
Antofagasta down to Magallanes. The reason to make this sub-division derives from the fact
that the North and Central-South Chile’s natural gas pipeline networks are not interconnected.
This sub-division also follows Chile’s electricity transmission network, which is made up of
four separate grids operating in the country (CNE, 2015): the Sistema Interconectado Central
(SIC) is the largest and serves the most populous areas, including the capital Santiago; the
Sistema Interconectado del Norte-Grande (SING) is the second largest and serves the
northernmost portion of the country, where most of the copper mining industry operates. For
this study the other two smaller grids, the Sistema Electrico de Aysen (SEA) and the Sistema
Electrico de Magallanes (SEM), were incorporated into the much larger SIC in the Central-
South Chile Region.
In order to divide the natural gas consumption in Chile in two regions (SING and SIC),
we used data from (SEC, 2015) (Table 10.1) to calibrate the base-year for each sector in each
region. Table 10.2 shows the method used to split the future increments of the national natural
gas demand for end-uses projected in Chile according to the drivers of the potential market
adopted.
189
Table 10.1.Sectorial natural gas consumption in Mm3 in Chile for 2006 and 2012.
Source: (SEC, 2015).
Interc
onnected
Systems SIGN SIC
Admin
istrative
Regions
X
V I II
I
II
I
V V RM
V
I
V
II
V
III IX
X
IV X
X
I XII
A
rica y
Parinacota
T
arapacá
An
tofagasta
A
tacama
C
oquimbo
V
alparaíso
Metr
opolitana de
Santiago
O
'Higgins
M
aule
B
ío-
Bío
L
a
Araucanía
L
os
Ríos
L
os
Lagos
A
ysén
M
agallanes
Sector
2012
Residential 0 0 1 0 0 25 229 0 0 9 0 0 0 0 183
Commercial
and Public 0 0 0 0 0 13 61 0 0
1
7 0 0 0 0 53
Transport
(CNG) 0 0 0 0 0 0 9 0 0 0 0 0 0 0 13
Industrial 0 0 0 0 0 132 375 31 0 0 0 0 0 0 92
2006
Residential 0 0 0 0 0 24 217 0 0 7 0 0 0 0
15
9
Commercial
and Public 0 0 0 0 0 10 40 0 0 3 0 0 0 0 44
Transport
(CNG) 0 0 0 0 0 1 21 0 0 0 0 0 0 0 11
Industrial 0 0 19 0 0 234 376 38 0 8 0 0 0 0 81
190
Table 10.2. Natural gas demand Split between Chile regions for base-year and
future increments of demand.
Base-year Demand Future increments of Demand
Sectors North Chile
Central-
South Chile Sectors
North
Chile
Central-
South Chile Drivers
Residential 0.3% 99.7% Residential 6.0% 94.0% Households
Commercial
and Public 0.1% 99.9%
Commercial
and Public 6.0% 94.0% Households
Transport
(CNG) 0.0% 100.0%
Transport
(CNG) 8.1% 91.9%
Number of
vehicles
Industrial 0.0% 100.0% Industrial 20.2% 79.8%
Fuel Oil
Consumption
Econometric approach for the natural gas demand projection in industrial sector
Table 10.3shows the parameters estimated for the ARIMA model constructed to
project natural gas consumption in the Argentinian industrial sector.
Table 10.3. ARIMA model for the natural gas consumption in the Argentinian
industrial sector.
Variable Coefficient t-Statistic
C 0.024268 [12.30524]
AR(1) 0.399686 [2.71118]
MA(1) -0.97369 [-39.48881]
Adjusted R-squared 0.286443
F-statistic 7.222148
Akaike info criterion -2.723897
Schwarz criterion -2.586484
Table 10.4 shows the parameters of the VECM approach to estimate natural gas in
the industrial sector of Bolivia.
191
Table 10.4. VEC model parameters for natural gas consumption in industrial sector
and GDP
Cointegrating
Equation: CointEq
Error
Correction: Δ(LnNGind) Δ(LnGDP)
LnNGind(t-1) 1 Cointeq -0.233 0.037
[-2.379] [ 3.2190]
LnGDP(t-1) -1.4046 Δ(LnNGind(t-
1)) -0.2007 -0.0337
[-6.922] [-1.3431] [-1.91691]
Δ(LnGDP(t-1)) 3.2617 0.2916
C -3.0348
[ 2.2274] [ 1.6928]
C2 0.0068 0.0234
[ 0.1489] [ 4.3088]
Akaike AIC -1.1022 -5.382
Schwarz SC -0.919 -5.1988
Note: t statistics in [ ]
In the case of Chile, in Figure 10.1is showed the shares by fuel of the historical
data of energy consumption in the industrial sector excluding electricity in Chile.
Figure 10.1. Energy consumption structure projected (from 2013 onwards) by
fuels excluding electricity in Chile’’s industrial sector.
Source: (IEA, 2014b)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Coal and Coke LPG Primary solid biofuels Diesel Natural Gas Fuel Oil
192
Figure 10.2 shows historical and projected values estimated for the useful energy-
excluding electricity- consumption in the Chilean industrial sector. The historical values
of useful energy were calculated based on the historical values of energy consumption
by fuels in the industrial sector from IEA (2014b), and efficiencies by fuel in the
industrial sector from (MIEMDNETN, 2009)22
.
Industrial GDP sector were estimated based on historical GDP from IMF (2015)
and % of Industry in the Chilean value added from WORLD BANK (2015). For future
industrial GDP we kept the same % of Industry in the Chilean Value added of 2012.
Figure 10.2. Forecasted useful energy-excluding electricity- in the Chilean
industrial sector. Source: (IEA, 2014b).
The two series, useful energy in the industrial sector-excluding electricity- and
industrial GDP showed to be non-stationary but integrated at first differences I(1),
Johansen Test Johansen’s test results showed cointegration between useful energy-
excluding electricity- consumption in industrial sector and industrial GDP. Based on
22Since there is no public data of efficiencies by fuel use in the Chilean industrial sector, we
adopted data from (MIEMDNETN, 2009) as a proxy.
0
1000
2000
3000
4000
5000
6000
7000
ktoe
Forecasted Useful Energy Estimated historical useful energy
193
the Akaike information criteria we chose two lag to construct a Vector Error Correction
model (VECM) for these two cointegrated series (Table 10.5).
Table 10.5. VEC model for useful energy consumption in industrial sector –excluding electricity-
and industrial GDP of Chile.
Cointegrating
Equation: CointEq
Error
Correction: Δ(LnNGind) Δ(LnGDP)
LnNGind(t-1) 1 Cointeq -1.983601 -1.27662
[-4.28236] [-2.84836]
LnGDP(t-1) 0.050625 Δ(LnNGind(t-1)) 0.392922 0.808967
[-6.922]
[ 1.19051] [ 2.53316]
Δ(LnNGind(t-2)) -0.030917 0.362661
[-0.16346] [ 1.98162]
Δ(LnGDP(t-1)) 0.522975 -0.391482
[ 2.83044] [-2.18972]
Δ(LnGDP(t-2)) 0.309218 -0.165317
[ 1.76229] [-0.97372]
C -0.048101 C2 0.0068 0.0234
[ 0.1489] [ 4.3088]
Akaike AIC -1.1022 -5.382
Schwarz SC -0.919 -5.1988
Note: t statistics in [ ]
10.2 ANNEX B: MODELLING SUPPLEMENTARY MATERIAL
ANNEX B presents additional data used for modelling power and upstream and
midstream technologies of natural gas.
Power Generation Technologies
The following Table shows the parameters and costs used to model the power
generation technologies. The Dual Combined Cycle and Dual Open Cycle power plants
corresponds to power plants installed in Chile as contingency to the interruption of
Argentinian natural gas exports in the last decade.
194
Table 10.6. Techno-Economic parameters for the power generation Technologies
modelling.
Source: BLACK&VEATCH (2012); IRENA (2015); PORTUGAL-PEREIRA et
al., [s.d.])
Technology Fuel Efficiency Investment Cost (US$/kW)
Fixed O&M Cost (US$2012/kW)
Variable O&M Cost (MUS$2012/PJ)
Lifetime (years)
Hydro Power Water 3500 15 0.72 40
Combined Cycle Natural Gas 0.455
1230 6.31 0.92
25 Diesel 0.435 6.3 0.92
Dual Combined Cycle
Natural Gas 0.435 1353 6.31 0.92 25
Diesel 0.415
Open Cycle Natural Gas 0.393 651 5.26 1.08 20
Diesel 0.245 651 5.3 2.56 20
Dual Open Cycle Natural Gas 0.373
716 5.3 2.56 20 Diesel 0.225
Diesel ICE Diesel 0.262 1110 32.9 3.33 5
Fuel Oil ICE Fuel Oil 0.338 1420 22 2.44 10
Biogas ICE Biogas 0.221 1300 22 2.44 10
Fuel Oil Steam Turbine Fuel Oil 0.369 1300 23 0.61 30
Coal Steam Turbine Coal 0.308 1300 23 0.61 30
Biomass Steam Turbine Biomass 0.115 1300 23 0.61 30
Solar Photovoltaic Solar 5300 50 0.00 20
Wind Power Wind 1810 50 0.00 20
Geothermal Geothermal 0.308 2720 31 0.00 30
Nuclear Uranium 6100 127 6.92 60
Concentration Solar Power Solar 5208 65 0.60 30
Multi-Hubbert parameters
The following tables shows the fitting parameters and EUR used for the Multi-Hubbert
oil production curve for Argentina and Brazil.
195
Table 10.7. Multi-Hubbert model parameters used for Argentina.
Source Tmax (Years) PM (MM3/year) b k EUR (Mm3)
H1 0.10 1.00 913
H2
0.24 1.00 416
H3
0.10 1.00 1080 Multi-Hubbert 2000 46.6 -
13 566
Table 10.8. Multi-Hubbert model parameters used for Brazil.
Scenario Source Tmax (Years) PM (kb/d) b k
EUR (Mm3)
Low Oil Production Scenario
Post-Salt (P95)
Onshore 0.15 1.00 1414
Offshore<400 m
0.16 0.94 2287
Offshore>400 m H1
0.19 1.00 11643
Offshore>400 m H2
0.06 1.00 13 566
Multi-Hubbert 2010 1957 – 28910
Pre-Salt (30)
Pre-Salt
0.13 1.00 30 000
000
Hubbert 2041
2 568 30 000
000
High Oil Production Scenario
Post-Salt (P50)
Onshore 0.15 1.00 1414
Offshore<400 m
0.16 0.94 2287
Offshore>400 m H1
0.22 1.00 6387
Offshore>400 m H2
0.09 1.00 37 695
Multi-Hubbert 2031 2362 – 47783
Pre-Salt (100)
Pre-Salt
0.14 1.00 100 000
000
Hubbert 2051
9 774 100 000
000
Gas-to-oil ratio for North Brazil
In Figure 10.3 a logarithm curve fitting was made on the gross associated gas produced
discounted by the re-injected gas over the oil production in the Amazonas State, in the
last 5 years. In the horizontal axis of this Figure year 1 represents the year 2010.
196
Figure 10.3. /oil
produced ratio equation adopted for the North Region.
Natural Gas reinjection in Oil Fields
The complexity and costs of implementing gas reinjection increase significantly
as it moves to ultra-deep waters. This is the case of Brazilian Pre-Salt Oil Fields. In the
case of Lula field, for instance, reinjection is being applied using a Waterflood,
Gasflood and Water-Alternating-Gas (WAG). In this process the gas can be a
hydrocarbon from the reservoir, or CO2 originally contained in the associated gas,
stripped in the FPSO processing plant (PIZARRO; BRANCO, 2012). In this case, the
cost for CO2 reinjection is estimated at 88 MUS$/Mm3 per day of capacity, based on
the well production costs23
.
Natural gas re-injection has different purposes than CO2 reinjection. In the second
case, CO2 is miscible with medium and light oil (above 28° API) at lower miscibility
pressures than natural gas (VERMA, 2015). In the case of natural gas reinjection using
the WAG technique, it acts as a secondary recovery method that can improve oil
23This high cost confirms why it is uncommon the practice of EOR recovery using CO2 in
offshore deep-water fields and in most of cases unprofitable without any CO2 taxation (KEMP, A. G.;
KASIM, 2013).
y = 0.0824ln(x) + 0.065
0
0.05
0.1
0.15
0.2
0.25
0 1 2 3 4 5 6
MMm3/Mbbl
Gas to oil ratio Logarítmica (Gas to oil ratio)
197
recovery by 5%, according to Rosa et al. (2016). In Brazil, wet natural gas re-injection
containing high CO2 contents is already occurring. However, in this case, the lack of
space in the FPSO to purify all the associated gas produced (SILVA, 2015) is the major
reason behind this fact. Somehow, this re-injected gas could be considered as a
contingent resources that in the future can become commercially feasible to be
extracted.
To consider the benefits of gas injection in the increase of oil production and
consequently in the revenues generation , a yield of 0.6 t of gas per barrel of EOR oil
was adopted based on the range between 0.38 and 0.63 t of CO2 per barrel of EOR oil
yields reported in KEMP and KASIM (2013). This barrel of oil produced from EOR
will be sold for a price of 60 US$/bbl in TIMES-ConoSur.
