See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271224965

Analysis of past and future oil production in Peru under a Hubbert approach

Article in Energy Policy · February 2015

DOI: 10.1016/j.enpol.2014.11.028

CITATIONS READS

18 183

3 authors:

Mauro F. Chavez-Rodriguez Alexandre Szklo

Federal University of Rio de Janeiro Federal University of Rio de Janeiro
22 PUBLICATIONS 348 CITATIONS 232 PUBLICATIONS 5,121 CITATIONS

SEE PROFILE SEE PROFILE

André F. P. Lucena
Federal University of Rio de Janeiro
108 PUBLICATIONS 2,452 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Forest conservation policy View project

Industrial Ecology and Biorefining: potential for promoting regional development in the Norte Fluminense region (RJ, Brazil) View project

All content following this page was uploaded by Alexandre Szklo on 28 May 2021.

The user has requested enhancement of the downloaded file.

 Energy Policy 77 (2015) 140–151

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

Analysis of past and future oil production in Peru under a Hubbert

approach
Mauro F. Chavez-Rodriguez n, Alexandre Szklo, Andre Frossard Pereira de Lucena
Energy and Environment Planning Program, COPPE/UFRJ – Universidade Federal do Rio de Janeiro Cidade Universitária, Centro de Tecnologia, Bloco C, sala
211, Ilha do Fundao, CEP: 21.945-970, Caixa Postal 68.501, Rio de Janeiro – RJ, Brazil

H I G H L I G H T S

Two scenarios of estimated ultimately recoverable resources were simulated.

A multi-cycle Hubbert approach better depicted the case of oil production in Peru.

Pipelines expansion and institutional and regulation changes explain this multi-cyle.

Institutional capacity and socio-environmental risks are barriers to oil activities.

art ic l e i nf o a b s t r a c t

Article history: This study evaluates scenarios for the oil production in Peru applying a Hubbert model. Two scenarios for
Received 2 September 2014 the estimated ultimate recovery (EUR) were proposed: the first, in which low investments in E&P and
Received in revised form social and environmental barriers undermine the development of oil resources beyond the limits char-
23 October 2014
acterized as 2P; the second, more optimistic, in which current exploratory and production areas in
Accepted 20 November 2014
Amazonia and low-explored Offshore-Shelf basins are developed, thus, increasing EUR to 3P reserves
Available online 1 December 2014
plus contingent resources. Findings show that oil production in Peru has not followed a Single-Hubbert
Keywords: pattern, except for the area with more drilling activity and the highest accumulated production in the
Oil production Northwest coast. Actually, institutional and regulation changes and less-attractive periods for operators
Peru
due to poor results in oil discoveries explain why a multi-Hubbert approach better depicted the oil
Hubbert model
production in Peru. Peru has the potential to achieve a second peak of 274 kbpd of crude oil, overcoming
Energy planning
Amazonia the peak of 195 kbpd, reached in 1982. However, most of the remaining production would be located in
Amazonia, where social and environmental issues pose critical challenges.
& 2014 Elsevier Ltd. All rights reserved.

1. Introduction In his classical study, Hubbert predicted in the late 1950s that
the production of oil in the continental United States would reach
Oil production forecasting techniques are usually divided into a peak in the beginning of the 1970s (Hubbert, 1956). Furthermore,
three main approaches: the economic, the geophysical based Hubbert (1959) related his graphical predictions for cumulative
models, and hybrid models, which combine the first two ap- production over time to a logistic curve. This behavior of oil dis-
proaches, aiming at explaining the deviations of the geophysical covery rate is explained by the combinations of two effects (Rehrl
models from the historical production (Benes et al., 2012; Kauf- and Friedrich, 2006):
mann, 1991; Reynolds and Baek, 2012). The economic models
stand on economic factors such as oil prices, costs, regulatory and a) Increasing information with exploration and cumulative dis-
technological issues to explain the evolution of oil supply (Fattouh, coveries, considering geological information and technical
2007). The geophysical approach is usually based on fitting oil know-how.
discoveries and production data models. For this latter approach, b) Decreasing recovery rate with exploitation and cumulative
the most applied models are based on the Hubbert's theory of oil discoveries, given the “depletion” of oil resources.
depletion.
For a price taker producer, trends in discovery should be di-
n
Corresponding author. rectly related to trends in production. Since operational adjust-
E-mail address: [email protected] (M.F. Chavez-Rodriguez). ments from producing oil fields are relatively inflexible, timely

http://dx.doi.org/10.1016/j.enpol.2014.11.028
0301-4215/& 2014 Elsevier Ltd. All rights reserved.
 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151 141

investments in new fields are necessary to increase oil production As many energy scenarios agreed that oil will remain a main
and/or replace declining fields (Rehrl and Friedrich, 2006). primary energy source worldwide and in Peru in the next years
In addition, the Hubbert cycles hypothesis is backed up by the (Alejos, 2011; IEA, 2013), forecasting oil production is crucial for
Central Limit Theorem, provided that mines/wells in the popula- energy planning and policy-making. Estimating future Peruvian oil
tion are numerous, with few regulatory constraints. Then, world- production and evaluating whether it will meet domestic demand
wide production can be treated as a sum of independent random is the first step to propose energy policies related to oil security
variables that in turn describe the production of individual wells/ (efficiency targets, fuel substitution, etc.) and allows evaluating
mines. By the Central Limit Theorem, the distribution of a sum future revenues from oil royalties. Projecting oil revenues, in turn,
tends to be normal or Gaussian (Patzek and Croft, 2010). is necessary to plan how to invest this government take.
Hubbert models are a widely used approach to predict future The last energy planning study published by the Government of
production of exhaustible natural resources, in particular of fossil re- Peru (MEF, 2011), forecasted two oil production scenarios up to
sources (Almeida and Silva, 2009; Brandt, 2010). Several authors have 2040 (Fig. 1). In the base case scenario, operators meet their
used Hubbert models (or their variants) to forecast world oil pro- minimum contract obligations and develop production at rates
duction (Bartlett, 2000; Brecha, 2012; Campbell and Laherrere, 1998; compatible with historical production. This slightly compensates
Gallagher, 2011; Hubbert, 1962; Maggio and Cacciola, 2009; Mohr and the natural declining of currently producing fields. The optimistic
Evans, 2010; Nashawi et al., 2010; Rehrl and Friedrich, 2006; Reynolds, scenario assumes an increase in investment in exploration and
2014; Wang et al., 2011); in addition, there have been efforts to fore- production of heavy oil in the Peruvian Amazon, hereinafter re-
cast oil production in some countries such as USA (Hubbert, 1956; ferred as “Selva”. This forecast was based on a production curve
Reynolds and Zhao, 2007), Former Soviet Union (Laherrere, 2002), that considers a declining rate of 7% per year plus the contribution
China (Tao and Li, 2007), Brazil (Saraiva et al., 2014; Szklo et al., 2007), of new oil fields and development activities on discovered fields.
Colombia (González et al., 2013a), North Sea Oil Fields (Blanchard, Since Peru is not a price-maker, such as some of OPEC members
2000) and OPEC countries (Nashawi et al., 2010), among others. (Ramcharran, 2002), it is worth testing if this country, in the same
The results of Brandt (2007), who tested Hubbert model for 139 way as non-OPEC countries, such as Brazil (Szklo et al., 2007),
oil producing regions, suggest that Hubbert's simple curve does would follow a Hubbert model for describing oil discovery and
not track very well the production rates of most regions. In production. It is also worth modeling the best Hubbert fit for Peru.
addition, Laherrere (1997) states that oil production should Given the lack of scientific peer-reviewed literature about the
be modeled with more than one Hubbert cycle (the so-called Peruvian oil industry, this study aims at filling this gap, by de-
“multi-Hubbert”). For instance, this approach can explain the veloping a Hubbert approach to model oil production in Peru,
production pattern in many countries with few basins and fields, initially by using a single cycle, and then by proposing three cycles
which have more than one peak in their production profiles (La- (multi-Hubbert). Furthermore, it contributes to energy planning
herrere, 1997; Nashawi et al., 2010). These multiple cycles are and policy-making processes related to the oil production in the
dependent on the expected profitability of each cycle's oil re- country.
sources, as it influences discovery and production (Rehrl and
Friedrich, 2006). Furthermore, under a country level analysis, de-
viation of oil production from a single-Hubbert pattern are sug- 2. Oil production in Peru
gested to be caused by institutional framework (Reynolds and
Kolodziej, 2007; Reynolds and Pippenger, 2010). Hence, multiple In 1863, the first oil well was drilled in the Talara basin,
cycles can be originated also from institutional and regulatory northwest coast of Peru, where light oil (35° API) was found. In-
changes (Reynolds and Zhao, 2007). Indeed, this paper provides terestingly, this was the first exploratory well drilled in South
evidence that a multi-Hubbert approach may be used to analyze America, just four years after the oil well drilled by Colonel Drake
the effects of institutional ruptures in oil production. in USA (Talara, 2006). Years later, the world's first giant field of La
Moreover, the simple Hubbert assumption is that oil produc- Brea-Parinas (1.14 Gbbl of oil) was discovered in this region
tion profile should be symmetric. However, evidences from the oil (Campbell, 2013). Fig. 2 shows the map of Peru and the basins with
production profiles of different countries suggest that these pro- hydrocarbon resources.
files are slightly asymmetric, with slower rates of decline than Peru was the biggest oil producer in South America until 1924,
rates of increase (Brandt, 2007). when Venezuela took the lead, followed by Colombia in 1927
In Peru, oil represents around 45% of the primary energy con- (Schlumberger, 2007). At the beginning of the 20th Century, oil activity
sumed (63% of which corresponds to imported oil). In 2010, Peru in Peru was controlled by private companies, particularly the “Inter-
net oil imports rose to near 27.8 MMbbl, resulting in a deficit of national Petroleum Co. Ltd.” (the Canadian subsidiary of the Standard
837.8 MMUS$ in the trade balance (MINEM, 2011). Oil of New Jersey), “Lobitos Oilfields Ltd.” and “Establecimiento

Fig. 1. Official scenarios for oil production in Peru. Source: Based on MEF (2011).
142 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151

Fig. 4. Wells drilled in offshore shelf and Selva regions from 1954 to 2012. Source:
Based on PERUPETRO (2013).

Although there are registries of offshore wells drilled at the

beginning of the 20th century in the Talara Basin, E&P activities in
offshore regions only intensified after the 1960s, mainly operated
by the American company Belco Petroleum Corporation (PER-
UPETRO, 2013) – see Fig. 4.
In 1965, Peru produced 62 thousand barrels per day (kbpd) of
oil and exported 3.3 kbpd; then, in 1969 EPF was replaced by the
national oil company “Petroperu S.A.” created by the D.L. No 17753
by nationalizing assets of the International Petroleum Co. Ltd.
(Touzett, 2007; Maurer, 2011).
According to Pontoni (1982), the decade of 1970s was char-
acterized by the intensive exploitation of the northwest coast. For
a century, this area accounted for most of the Peruvian oil pro-
duction, totaling 96% of the country's accumulated oil production
Fig. 2. Map of basins and wells drilled in Peru at August of 2012. Source: Based on at the end of 1976 (978.5 million barrels). In addition, almost all
PERUPETRO (2013) and USGS (2000). the oil resources exploited until the 1970s were already known in
the 19th century (Schlumberger, 2007).
In spite of the 4 unsuccessful new field wildcats made by
Texaco in the Marañon Basin (Amazon) in the 1950s, in 1971 the
Corrientes Field was discovered from Petroperu's first well in the
region, followed by Oxy's discovery of the Capahuari field in 1972
(Perupetro, 2002).
By 1977, due to the high growth rates of oil consumption, along
with a slow decline of the northwest coast fields, Peru imported
41 kbpd, even though production had risen to 91 kbpd. However,
owing to the development of important oil fields discovered in the
North Selva (Marañon Basin), Peruvian oil production regained.
Selva became the major producing area of Peru, with the support
of the strong oil policy of the early 1970s. This was characterized
by the opening of the Amazon and the completion of the Northern
Peru Oil Pipeline (Transandean) in 1977, first assigned to PETRO-
Fig. 3. Liquid hydrocarbon production in Peru and oil prices. Source: Based on BP
PERU in 1972 (PETROPERU, 2013). Between 1971 and 1978, 261
(2013), Campbell (2013), MINEM (2000, 2013a), Pontoni (1982) and Vásquez
(2005). wells were drilled in Peru, 49% of which in the North Selva. During
this period proven oil reserves increased from 332 MMbbl to 720
Industrial de Petróleo Zorritos”. In 1907 these three companies pro- MMbbl (Pontoni, 1982). By February of 1978 domestic production
duced around one hundred thousand tonnes of oil (2 kbpd) (Calmet, again exceeded consumption.
2012); in 1921, the country production was 10 kbpd, increasing to Therefore, although the increases in prices in the 1970s led to a
48 kbpd in 1936 before decreasing to 32 kbpd in 1941 (see Fig. 3). In higher production, new discoveries and developments in oil
this period, in spite of low oil prices, the government pressured the transport infrastructure that were already underway before the
companies to increase their production in order to match its own fiscal first oil shock in 1973 also created the conditions for that increase
demands. This resulted in a reduction of efforts in exploration but to happen. Furthermore, during the military government of that
intensified exploitation in discovered fields. time, new concession contracts with foreign companies were
Exploration investments in the Amazon Region began in 1939 made, which created the institutional framework for that. In 1981,
when the company Mobil discovered the Aguas Calientes Field in exports of crude oil and oil products accounted for 21% of Peru's
the Ucayali river basin (Vásquez, 2005). In 1948 the first national total export (BCRP, 2012). That year, oil production reached
oil company (NOC) was created: “Empresa Petrolera Fiscal” (EPF). 193 kbpd and the Selva region represented 64.2% of total pro-
However, private companies kept their predominance in Peruvian duction. The main producer was the private company Occidental
oil activities. in Selva (105 kbpd) followed by the NOC Petroperu (42 kbpd).
 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151 143

Since 1982, oil production began to decrease. Some discoveries

of small non-commercial oil accumulations, mostly heavy crude oil
were made in the Amazon basin (Sanchez-Sierra et al., 1984), but
except for the Camisea natural gas fields discovered in 1985 there
were no other major discoveries in Peru. Investments also de-
creased significantly due to the few encouraging exploration effort
results achieved (Vásquez, 2005), the negative impacts of eco-
nomic policies – especially those related to Petroperu –,1 low oil
prices, and the presence of insurgent groups in hinterlands after
the 1980s (Quijandria, 1993).
Then, the new administration in the early 1990s established
important reforms for the hydrocarbons industry, including the
adjustment of price distortions and privatization of the down-
stream and upstream business of Petroperu (2012). A new law for
the hydrocarbons sector was passed through in 19932 aiming at
encouraging private investments in oil and gas activities. It es- Fig. 5. Evolution of reserves in Peru 1990–2012. Source: based on MINEM (2013a,
tablished that hydrocarbons resources in the ground belong to the 2005, 2000).

State but, once extracted, it belongs to the producer. This meant

free disposal of the hydrocarbons produced and free import/export
4Q m 5
of crude oil and products, prices set by supply/demand forces, and b= =
U c (2)
free entrance of private companies to any activity in the oil/gas
industry not mandatorily in association with Petroperu (ESMAP,
EUR = 0.8Q mc (3)
1999). As a result, from 1995 to 1998 foreign investments in ex-
ploration rose significantly, from 36.7 MMUS$ to 229.4 MMUS$. In which EUR is the estimated ultimate recovery (or ultimate
However, there were no significant crude oil discoveries in Peru, recoverable reserve), Q is oil production at time t, Qm peak pro-
which compromised the investment in oil exploration in the fol- duction, tm the time of the peak, b and c are parameters which
lowing years. account for the slope of the curve and the estimated average useful
In 2003, investments in exploration were as low as 12 MMUS$. life of fields, respectively.
However this downward trend was reverted during the 2000s, and The definition of EUR is fundamental for the Hubbert model
in 2012 investment in exploration reached 958 MMUS$, mainly (Guseo et al., 2007; González et al., 2013b), since defines the peak
driven by the exploration in Selva region (MINEM, 2013a, 2005). production, the time of the peak and total oil produced under the
Moreover, between 2003 and 2012 investment in exploitation rose Hubbert curve. Bentley et al. (2007) indicate that the most accu-
from 347 MMUS$ to 921 MMUS$, owning partially to the sig- rate data on reserves should be associated with the probability of
nificant investments made by Pluspetrol Perú Corporation S.A. in proven plus probable reserves – or reserves 2P (where P relates to
the Camisea Project (blocks 56 and 88) and Pluspetrol Norte S.A. in the probability of reserve addition – SPE, 2007). Owen et al. (2010)
block 1-AB and 8 in north Selva. Liquid hydrocarbons output in- agree with this finding, since “assuming estimates are accurate, 1P
reserves would be expected to be revised upwards over time and
creased from 91 kbpd in 2003 to 153 kbpd in 2012, driven by the
3P reserves downwards to converge at the estimated 2P volume.
increasing production of liquids from natural gas, as crude oil
For this reason, 2P reporting should represent actual reserve vo-
production has decreased (MINEM, 2005, 2013a).
lumes most accurately”. Fig. 5 shows the evolution of reserves in
Finally, in the offshore areas of Peru hydrocarbon exploration
Peru in the two last decades, in which 3P reserves have varied
and production activities has been concentrated in shallow waters
significantly when compared to 2P reserves. Proven reserves (1P)
(depths of less than 100 m). In the northern part of the country,
have remained stable.
exploration has gone to depth of no more than 120 m. According
This study proposes a conservative EUR scenario (EUR1), which
to Perupetro (2011) it is expected that E&P activities will rise, includes the 2P reserves of Northweast Coast, Offshore Shelf and
especially for heavy oils in the northern jungle, and natural gas Selva. This EUR scenario assumes a reserve growth based mostly
and liquids of natural gas in Camisea and the surrounding areas. on discovered fields. This agrees with the “best estimate” EUR for
forecasting in other works such as Owen et al. (2010), Szklo et al.
(2007), Saraiva et al. (2014), Campbell and Laherrere (1998),
3. Material and methods among others. For the accounting of 2P reserves is remarkable the
contribution of Block X (former owned by Petrobras and recently
3.1. Forecasting Peru's oil production according to a single Hubbert transferred to CNPC) in the Coast with 171.3 MMbbl, Blocks Z-2B
model (single cycle) (Savia) and Z-1 (BPZ) with 112.8 MMbbl and 97.7 MMbbl, respec-
tively, in Offshore Shelf, and Blocks 1AB (Pluspetrol Norte) and 67
The Hubbert model used in this paper is described by the fol- (Perenco) with 177.8 and 219.0 MMbbl, respectively, and Block 39
lowing equations (Moroney and Berg, 1999) – see Eqs. (1)–(3): (Repsol) with 273.9 MMbbl currently in exploration stage in Selva
– see PERUPETRO (2014a) to view the spatial distribution of the
2Q m Blocks.
Q=
1 + cosh[ −b(t − tm)] (1) In order to establish a more optimistic forecast, this study also
defined the EUR2 scenario, which is composed by the sum of 3P
reserves and total resources adjusted by the ratio between fractile
1
The fuel pricing policies (including taxation) applied by different govern- F95 and fractile F5 provided by USGS (2012) for undiscovered oil in
ments to alleviate fiscal deficit or contain inflation, drained resources from the
company, lowering its capacity to suitably invest in E&P activities (Hamann and
Peru. This corresponds to a EUR2 91% higher than EUR1, which is
Paredes, 1991; Campodónico, 1999). also higher than the optimistic forecasts of other authors such as
2
Organic Hydrocarbon Law, No. 26221, which is currently in force. Brecha (2012) or Szklo et al. (2007). The evaluation of the EUR2
144 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151

scenario intends to forecast an upper limit for future oil produc- adopted (Table A.2). RMSE is defined as follows (Nashawi et al.,
tion in Peru. For example, the optimistic forecast of Saraiva et al. 2010):
(2014) for Brazil used P5 reserves, resulting in a EUR correspond-
ing to 123% higher than the P50 scenario. ∑in= 1 (Q act − Q cal )2
RMSE =
In terms of Possible Reserves, the Selva Region has the highest n (4)
participation, with remarkable contribution of Blocks 1AB with
where Qact and Qcal represent the actual and calculated production
91 MMbbl, and Block 67 with 123 MMbbl. Both these blocks are in
rates, respectively, and n is the number of data points.
production. Block 39 (Repsol) and 64 (transferred to Petroperu by
Talisman, who decided to withdraw operations in 2012) with
104 MMbbl and 158 MMbbl, respectively, are both under ex-
3.2. Forecasting Peru's oil production according to a multi-Hubbert
ploration. Finally in terms of resources, besides the blocks under
model
development or under exploratory contract (where Offshore Shelf
blocks stand out), the participation of “Exploratory assessment”
According to Reynolds and Kolodziej (2007) and Reynolds and
areas – where authorization and licenses for exploratory activities
Pippenger (2010) interruptions caused by institutional and
are being launched – and the so-called “Non-Operated Areas” are
regulatory changes in oil production can cause deviations of oil
relevant.
production from a single-Hubbert pattern. Since the Peruvian oil
As resources in contracts for “exploratory assessment” are ac-
industry has passed through various institutional changes during
counted under a national level, this study assumed an allocation
its history, and recently some projects in the Northern Peru be-
among the areas according to the distribution of Possible Reserves
came profitable due to investments in oil pipelines, this study also
and Resources, respectively involved in E&P Contracts. “Non-Op-
performed a multi-Hubbert model for Peru's oil production, ac-
erated” areas are related to blocks located in Salaverry-Trujillo,
cording to Eqs. (5) and (6):
Lima, Lima-Pisco, Mollendo, Moquegua and Titicaca. Considering
that Moquegua and Titicaca basins do not possess significant re- N
2Q mi
sources and the remaining are offshore basins, resources in “Non- P= ∑
i=1
1 + cosh[ −bi (t − tmi)] (5)
Operated” areas were allocated in the Offshore Shelf. Finally, it is
worth mentioning that no shale oil prospects are yet known in N
Peru (USGS, 2013). Table 1 shows the basic information collected EUR = ∑ EURi
to build the single Hubbert model for Peru. i=1 (6)
This study applied two fitting procedures for the single-Hub-
where i represent each cycle, and N the total number of cycles
bert. The first approach fixes the EUR and then parameters b and tm
modeled.
are obtained by minimizing the root-mean-square of the errors
As stated by Anderson and Conder (2011), the precision to
(RMSE) in relation to the historical production until the peak-year
which historical production data can be fitted using a multi-
of production. The parameter b was limited to 0.27%/year, which is
Hubbert approach is dependent on the number of cycles used to
the highest slope of the multiple logistic curves used by Brecha
model the data. This study modeled three Hubbert cycles: two of
(2012). This limit value for b means an estimated average useful
them (H1, H2) were used to describe the historic production, as
life of fields (c parameter) of 18.5 years (Table A.1). The second
described in Section 2: H1 represents a normal growth of pro-
approach fixes both EUR and b, and estimates tm by minimizing
duction whilst H2 represents a rapid short term growth. Then, the
RMSE. For the Coast region parameter b of Eq. (1) was set at 0.058,
third cycle was used to model the future production, given the
following the results of Maggio and Cacciola (2009), which re-
remaining EUR in Peru. This cycle also helps to improve production
present an average useful life of fields worldwide. In turn, for
fitting of the last years.
Offshore Shelf and Selva, which rose and fell in a very short period,
This study set parameters b according to Table A.3, Qmi and tmi
a value of b equal to 0.1, as proposed by Brecha (2012), was
resulted from the minimizing of RMS for historical production.
Therefore, although the third cycle helped to improve the fitting of
the model, its purpose focuses more on projecting the future
Table 1 production curves for Peru than improving the understanding of
Reserves, accumulated production statistics and EURs for Peru. Source: Based on
MINEM (2013a, 2013b).
past ruptures. Clearly future ruptures that can create other cycles
may happen, but, for the sake of simplicity, this analysis tried to
Oil statistics Coast Offshore Selva Country depict boundaries for Peruvian oil production.
shelf

Accumulated production
By areas at the end of 2012 (MBbl) 1190 355 1038 2583 4. Results
Reserves and resources involved in
E&P contracts Fig. 6 depicts the results of the single-Hubbert model for each
Proven reserves 233 130 269 633 Peruvian oil producing region. Fig. 7 sums up these curves' results
Probable reserves 59 88 521 668 forming the country's total oil production curve.
Possible reserves 76 116 578 770 Findings show that oil production in Peru has not followed (or
Resources 347 591 699 1636
should not be modeled by) a single-Hubbert pattern. This con-
Under contracts for exploratory clusion is valid for the two fitting approaches undertaken in this
assessment
study (Fig. 7). Only for the Northwest Coast the single-Hubbert
Resources 1236
curve resulted in a relatively good fit with observed values, when
In non-operated areas
considering a conservative EUR scenario (EUR1). This can be ex-
Resources 1111
plained by the regular and historical investments made in this area
EUR for Hubbert modeling over the time. This is the area with the highest accumulated
EUR 1 1483 574 1828 3885
EUR 2 2030 2022 3356 7408
production and the highest numbers of relative small production
wells. According to estimates based on the data of PERUPETRO
 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151 145

Fig. 6. Single-Hubbert curves for Peru oil production by areas for two scenarios of EUR fixing EUR (a) or both b and EUR (b).

(2013) and USGS (2000), traditional basins in the Northwest Coast and 64 in the Selva region, which are currently at the exploration and
reach an exploration Intensity indicator of around 18 km2/ExpW development stage. This sensitivity scenario would reflect a con-
(square kilometer of the sedimentary basin/exploratory well servationist governmental policy to preserve ecological system and
drilled in the sedimentary basin). This indicates an exploratory indigenous communities from E&P oil activities in Amazonia (Haselip,
effort much higher than the one found in other basins, such as 2011). The results for this sensitivity scenario show that the current
Talara, Sechura and El Progreso basins (in the Offshore Shelf),
level of oil production could be maintained until 2023, after reaching a
whose exploration indicator reached 943 km2/ExpW; or Selva
peak of just 71 kbpd.
(including the Marañon and Ucayali basins), whose indicator is
On the other hand, results for an optimistic scenario based on
1776 km2/ExpW.3
EUR2 (Fig. 10) shows that Peru could overcome its historical peak
As the single Hubbert model did not fit very well with the
Peruvian oil production, this study undertook a multi-Hubbert of 1980 (195.3 kbpd) by 2032, and could reach a peak production
approach by areas. The results for the conservative EUR1 scenario of around 274 kbpd by 2046. Findings show that Amazonia would
are presented in Fig. 8. still be the main contributor to the oil production in Peru, but
Hence, the forecasts according to the conservative scenario of EUR1 Offshore Shelf would also become relevant, as well as the ex-
(Fig. 9) show that Peru would reach a peak oil production of around ploitation of the hydrocarbon resources in the Salaverry, Trujillo,
90.4 kbpd by 2029, similar to the historical value for 2003. Under this Lima, Pisco and Mollendo basins. The Northwest Coast does not
scenario, Amazonia remains the main contributor to oil production in show encouraging results. Even under the optimistic EUR2, pro-
the country, followed by the Offshore Shelf and the Northwest Coast, duction in this region would not overcome its historical peak,
which show a sustained declining, however. Moreover, this study si- reaching a local peak in 2031 followed by a slow declining, much
mulated a sensitivity scenario for excluding the reserves of blocks 39 like a double single-Hubbert. Parameters for MH1 and MH2 do not
vary significantly when the Multi-Hubbert model is tested for
3
All estimates were made within the Peruvian territory borders of each basin. EUR2, but EUR2 does cause a great impact on MH3.
146 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151

Fig. 7. Single-Hubbert curve for Peru oil production for two scenarios of EURs fixing EUR (a) or both b and EUR (b).

5. Discussions the Northern Amazon, resulting in another Hubbert-cycle. This

coincides with the years from 1968 to 1979 when the military
Findings have shown that the single-Hubbert approach should government subscribed concession contracts for foreign oil com-
not be used to model the Peruvian oil production, given the sig- panies, giving them a share of the produced oil under fiscal
nificant deviations from observed production curves. The analysis incentives as retribution for their exploration and production op-
of the causes behind this misfit can provide relevant information erations (Pontoni, 1982). These fiscal incentives, along with back-
and is an useful tool for policy makers. Actually, this procedure wardation expectation on oil prices and high interest rates created
was followed by Kaufmann (1991), Reynolds and Kolodziej (2007) conditions for optimal production capacity at high levels, accel-
and Reynolds and Baek (2012), in order to understand how vari- erating the depletion in detriment to recovery factors (Nystad,
ables such as oil price and technological and institutional factors, 1988, 1987). Then, on December of 1979, the government re-
would explain the deviations in the actual oil production curves negotiated the active contracts to include corporate taxes (Pontoni,
from the one based on the classic Hubbert model (1956). 1982). However no significant prospects were developed to com-
Modeling the Peruvian case indicates relevant deviations be- pensate the rapid declines.
tween observed figures and the production estimates based on In the 2000s, the Peruvian economy has grown at an average of
Hubbert. This happens mostly after 1980, along with low prices 6% per year (INEI, 2014). Crude oil and oil products imports ac-
(oil counter-shock in 1986), and changes in the regulatory fra- counted for 6084 MMUS$ (14.47% of total imports), while exports
mework encompassing the activities of oil operators (Pontoni, of these goods reached 3036 MMUS$ (7.25% of total exports), re-
1982). For instance, one observed effect of these events was the sulting in a negative trade balance of 3047 MMUS$.4
decrease in the number of the offshore wells drilled (Fig. 4). Since demand for fossil fuels is currently pulled by the eco-
Moreover, during this time, in the Selva region, there were no nomic growth the petroleum trade balance deficit will probably
significant discoveries (Vásquez, 2005). increase in a conservative scenario for the oil production, such as
On the other hand, the Multi-Hubbert model provided a better EUR1, which is more likely to occur. Consequently, the total net
fit to the historical oil production in Peru. Interestingly, the Multi- trade balance and the economy growth of Peru could be jeo-
Hubbert approach helped in identifying periods of accelerated pardized by the import of crude oil and oil products.
production, which occurred in the Offshore Shelf and Selva regions This perspective is also backed by the fact that no relevant
in the 1970s. During that decade, high oil prices resulted in large shale or tight oil prospects are yet known in Peru. In 2009, Maple
inflow of recycled petrodollars (El-Gamal and Jaffe, 2010), partially Energy found gas in the Devonian shale formation while drilling
diverted towards lending to developing countries and used by the Santa Rosa 1X well. The Devonian shale gas deposits were
large financial institutions to finance investments in oil frontiers encountered at a depths of 3909 m to 3976 m and 3978 m to
(Chevalier, 2004). In addition, during this time there was also an 3986 m, but no oil accumulations were found (Maple, 2012). After
active role of Petroperu in upstream and midstream activities.
Besides, in the 1970s, the Northern Peru Oil Pipeline (Trans- 4
This figure does not take into account Liquefied Natural Gas exportations
andean) was completed allowing the production expansion in (1372 MMUS$) (SUNAT, 2014).
 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151 147

Fig. 9. Multi-Hubbert curve for Peru oil production for EUR 1.

Fig. 10. Multi-Hubbert curve for Peru oil production for EUR 2.

under “force majeure clauses”, due to either the delay for ap-
proving environmental studies or to social conflicts (Gestión,
2014).
As the findings of this study indicate, the exploitation of Off-
shore Shelf hydrocarbon resources is crucial to increase the crude
oil production in Peru. However difficulties have appeared to
Fig. 8. Multi-Hubbert curves (MH) for Peru oil production by areas, for EUR1. promote the exploitation in these areas. For example, in 2013,
PERUPETRO, Peru's hydrocarbon regulatory agency, called for an
International Bidding Round for nine offshore blocks. Due to the
reviewing third party reports, the company concluded that, “due lack of knowledge about the hydrocarbon potential of these blocks,
to a variety of factors, there would likely be no interest from third only one company declared formal interest to participate, and,
party joint venture partners in further developing the shale gas thus, the bidding was postponed (Gestión, 2013). Therefore, efforts
discovery at this time and that the various potential development to develop seismic studies in these areas are to be put in place to
scenarios did not have attractive economics at this time under inform potential bidders. In 2014, after the conclusion of these
current market conditions” (Maple, 2013, p. 72). In 2013, Maple efforts, another bidding round for these offshore blocks is ex-
Energy reported to Perupetro S.A. that it had decided to cease pected to happen.
activities in Block 31-E related to the shale gas prospect (Maple, Nevertheless, to perform an effective promotion of oil activities,
2014). PERUPETRO's resources and capacities should be strengthened. For
Current conditions does not seem encouraging for petroleum example, the analogous entity of PERUPETRO in Brazil, the Brazi-
activities in Peru. At the end of 2010, 85 contracts were in force: 66 lian National Agency of Petroleum, Natural Gas and Biofuels (ANP),
exploratory and 19 exploitation contracts (PERUPETRO, 2014b). By is developing stratigraphic wells for Brazil's new frontier basins, in
April 2014 this number declined to 73 (49 exploratory and 24 addition to the solely acquisition of geophysical and geological
exploitation contracts). Nevertheless, 30 of the 73 contracts are (ANP, 2011).
148 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151

Improving data collection is the first step for understanding the reserves are small compared to other blocks in other countries,
challenges related to social and environmental threats and asses- they are located in areas of high biological diversity and are sig-
sing Peruvian resources (e.g. through mapping sedimentary ba- nificant enough to provide relatively large revenue streams for
sins). Even for the immature plays with wildcats drilled, such as Peru.
some blocks in the Amazon and offshore, knowing the resources Further research is needed to tackle the socio-environmental
and the extraction costs is fundamental for investors to determine conflicts. In some cases, grievances are exacerbated by a percep-
the margin of the prospects and, consequently, to assess the value tion that local communities receive a disproportionately low share
of the option of drilling again (Smith, 2005). of the benefits from exploration and extraction, but are expected
In addition, the country's institutional framework should be to bear the brunt of the social and environmental costs (Haselip,
able to attract operators, considering that Peruvian basins are 2011). In this sense, studies about compensation mechanism to
frontier areas that compete with prospects from other countries in local communities should also be conducted. Evaluating how to
Latin America (e.g. Mexico) and even Africa. A possibility is to spend the royalties in a sustainable way is important, since there is
establish partnerships with NOCs from emerging countries, espe- a local economic dependence on oil activities and there is a long-
cially those from neighboring countries, which may be attracted term risk of extreme poverty when oil runs out.
by logistic advantages and regional integration. Moreover, royalties could also be spent in environmental pro-
Following Nystad (1985, 1988), Peru can manage taxes to tection in order to overcome, at least partially, the environmental
increase the optimum recovery factor. Peru's regulatory frame- impacts of oil activities. Still, for some cases, such as those related
work is apparently considering such a tax incentive. For instance, to isolated indigenous communities, economic compensation is
hydrocarbon law provides that exploration and development not suitable, and other protection mechanism have to be proposed.
expenditures (even those in contract area not at commercial Taking into consideration delays in approving environmental
extraction stage), including the investments made before com- studies and other institutional barriers, SPH (2014) suggested
mercial production, can be accumulated in an account for further some measures to accelerate the approval of licenses for petro-
amortization. leum activities. Nevertheless, they should not weaken the Gov-
Hydrocarbons resources exploitation in Amazonia also poses ernment capacity to deal with the environmental impacts and
critical challenges. Scientific studies report that oil activities have risks related to these activities.
had significant deleterious effects on aquatic organism, impacts on Finally, some limitations of the methodology and results of this
indigenous communities health trough the exposure to toxic study should be highlighted. One of them is the definition of EUR.
substances, and impacts on the environment by spilling of waste EUR are not a category of resources, but a concept that can be
applied to any accumulation or group of accumulations, to define
and muds, petroleum spills and production waters (Reátegui-Zir-
quantities of oil to be potentially recovered under feasible tech-
ena et al., 2012; Martínez et al., 2007). Moreover, some impacts
nical and commercial conditions (González et al., 2013b). Some
might be insurmountable, especially those related to isolated in-
authors consider an EUR fixed according to geology, but others
digenous communities (Napolitano and Ryan, 2007).
introduce additional parameters such as knowledge growth,
Actually, there are strong linkages between environmental
technological advance and economic changes in the estimation of
impacts and social conflicts. By April 2014, social conflicts in Peru
EUR. Some economists avoid using the concept as a static value,
totaled 212, 136 of which were related to environmental issues. All
since the recoverable capacity depends on unpredictable economic
the 17 social conflicts in the Peruvian hydrocarbon industry were
and technological changes (Brandt, 2010; Reynolds and Baek,
related to socio-environmental issues (Defensoría del Pueblo,
2012; Sorrell and Speirs, 2009).
2014).
Other limitations include the assumptions of the Hubbert ap-
The indigenous activism in Peru has also increased in the last
proach. One is the assumption that the pace of oil production will
years. In 2011, the Law of Prior Consultation of Indigenous Peoples
follow the incorporation of reserves with a relatively constant time
(Law No. 29785) was enacted.5 However, the result of the con-
lag. Institutional factors, oil services local capacity, price variations,
sultation process are not binding and further social conflicts will
lack of transport infrastructure as well as bureaucracy delays and
likely occur, resulting in reputational and delay risks for investors.
social conflicts could distort this lag along the time affecting the
In this sense, this study performed a sensitivity analysis for
bell-shaped curve. Indeed, this is one of the major reasons behind
evaluating the cases with and without the production of blocks 39
some of the cycles defined here for the multi-Hubbert analysis of
and 64. For both blocks international oil companies have declined
Peru. Therefore, additional cycles can emerge from future in-
participation. For instance ConocoPhilips withdrew its participa- stitutional discontinuities of the Peruvian oil policy.
tion from block 39, and Talisman and Hess transferred to the state More detailed models for forecasting oil production, such as
oil company Petroperu the ownership of block 64. The block 39 bottom up and econometric models, are better for accurate near-
overlaps a national protected area and a proposed reserve for term projections. Nevertheless, they have also their shortcomings.
uncontacted indigenous groups (Orta-Martínez and Finer, 2010), For bottom-up models, detailed data at field level is needed, which
while, in the block 64, the former operators, Talisman and Hess, also means many more assumptions than requested by a typical
had to deal with conflicts with Achuar indigenous people. Cur- Hubbert model. On the other hand, although not always intensive
rently members of the Achuar have been protesting against Pet- in data collection, the forecasts based on econometric models
roperu plans to enter their territory and exploit oil resources as usually perform poorly for predicting far beyond the fitted data.
well (Amazon Watch, 2012; The Guardian, 2013). Econometric models are subject to a large number of omitted
The sensitivity analysis of this study indicated that, should variables (for which data are likely not available), which can affect
these two blocks not be exploited, a crude oil peak production of the values estimated for the models' coefficients (Brandt, 2010).
71 kbpd would be reached by 2023. Otherwise, a 90 kbpd oil As the purpose of this study is not to establish with a accurate
production would be reached by 2029. This is the direct effect of precision the future Peruvian oil production trajectory, but to de-
giving up around 2P reserves of 300 MMbbl. Although these fine boundary scenarios to support energy policies, a multi-Hub-
bert approach seems to be suitable. Indeed, according to Brandt
5
The Prior Consultation of Indigenous Peoples Law requires that licenses for oil
(2010) these type of Hubbert-based models are good enough to
exploration and production activities near indigenous communities have to go forecast the decade of the oil peak production for an estimate of
through a public consultation process. EUR in a particular region.
 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151 149

Findings also corroborate the conclusions of Watkins (1992) environmental compensation, income taxes may have to be
about the inadequacy of simple economic models based on the lowered.
Hotelling principle to cope with “petroleum industry realities”. Even though, there may remain insurmountable sources of
The case of Peru indicates that applying a Multi-Hubbert approach conflict in some Peruvian promising basins, whose analysis is out
may be justified not only by the expansion of EUR, which was also of the scope of this study, but should be undertaken in the future.
commented by Watkins (1992) as one of the drawbacks of Ho- In addition, future studies should also be carried on to devise
telling's principle, but also by institutional and even contractual or new roles of the Peruvian NOC, Petroperu, to promote upstream
logistic drivers. activities in the country. For instance, this might include partici-
pating in lesser profitable projects (marginal and even mature
fields), which are not attracting IOCs, as of today, and even develop
6. Conclusions and policy implications prospects with smaller rates of return.
Finally, should the more conservative EUR scenario happen
Peru was the first oil production country in South America, and together with a rising demand for oil products, the hydrocarbons
one of the largest oil producers in the region until the 1980s. Since trade balance deficit will increase, struggling the Peruvian econ-
1982, though, oil production began to decrease. Discoveries of omy development in the future. Current Peruvian energy policies
small non-commercial hydrocarbon accumulations, except for the try to address this issue, for example, by promoting fuel switch
Camisea natural gas fields discovered in 1985, did not reverse this between GLP and natural gas in the residential sector, and liquid
trend. fuels and natural gas, in the transport sector (MINEM, 2013c).
The case of Peru evidences that the oil production in some Nevertheless, further studies are also urgent to evaluate the im-
countries deviates from a single-Hubbert pattern, but can be quite pacts of these policies.
well modeled through multi-cycles. In Peru, these cycles were
mainly impacted by an accelerated expansion of the oil production
in the Selva region after the conclusion of the Northern Peru Oil Acknowledgments
Pipeline (Transandean), and the series of regulatory framework
changes along two decades that affected the attractiveness of We acknowledge the financial support from CNPq, particularly
upstream activities in Peru, a risky and marginal oil frontier area. the CNPq/TWAS (Processes 190318/2011-2). We also thank Ro-
To assess the oil peak level in Peru, through different EURs, berto Schaeffer for his contribution to the early stages of this work
including sensitivity analyses to deal, at least partially, with socio- and Alfredo Dammert for the revision and comments to this paper.
environmental issues, is relevant for the energy planning of the Finally, we thank Rolando Bolaños from PERUPETRO and Arild
country. This is particularly true for the decision making process Nystad.
related to energy security, investments in upstream and refining,
and diversification of the energy supply. Results showed that oil
production in Peru has the potential to achieve a second peak of Appendix A
274 kbpd6 by 2046.
Interestingly, although based on an optimistic scenario, this See Tables A.1–A.3.
production level is lower than the oil production of neighbor
countries, such as Colombia (944 kbpd) and Ecuador (502 kbpd) in
2012 (ANH, 2012; ARCH, 2012). In addition, the achievement of Table A.1
such scenario will rely on the exploitation of oil resources in the Parameters estimated for the single-Hubbert model fixing EUR.
offshore shelf and the Amazon basins. This is highly uncertain and
would require at least an attractive institutional framework to try Parameters EUR 1 EUR 2

to overcome social and environmental conflicts in areas with Coast Offshore Selva Coast Offshore Selva
promising oil resources. Hence, one major task should be shelf shelf
strengthening PERUPETRO, the hydrocarbon regulatory agency, as
happened in Brazil with ANP, for example. In fact, the Peruvian c (years) 91 20 19 104 21 19
EUR (MBbl) 1483a 574a 1828a 2030a 2022a 3356a
highest production potential is located in Amazonia, where the
Qm (MBbl/year) 20 37 123 24 119 226
social and environmental regulation of E&P activities must be b (%/year) 0.055 0.255 0.270 0.048 0.235 0.270
strengthened as well, emphasizing the protection of isolated in- tm (year) 1965 1982 1989 1975 1989 1991
digenous communities (Napolitano and Ryan, 2007). a
Fixed parameters.
Improving PERUPETRO's budget and capacity for data collection
and processing is important. In addition, since Peru is a risky
frontier production area, its institutional framework could
strengthen the role of the country's NOC in the upstream. One
possibility is to explore possible partnerships with other NOCs that Table A.2
either wish to add reserves or have financial resources to bear Parameters estimated for the single-Hubbert model fixing EUR and b.
greater risks. This includes, for example, South American and
Parameters EUR 1 EUR 2
Asian companies, and might also comprise investments in pipe-
lines and terminals. Finally, the Peruvian royalty charges are al- Coast Offshore Selva Coast Offshore Selva
ready high compared to some countries of Latin America (CEPAL, shelf shelf
2013). Increasing them may be justifiable from the social and en-
c (years) 86a 50a 50a 86a 50a 50a
vironmental point of view, but might undermine the country's
EUR (MBbl) 1483a 574a 1828a 2030a 2022a 3356a
production and, consequently, the revenues from it, In this case, in Qm (MBbl/year) 22a 14a 46a 30a 51a 84a
order to maintain the royalty revenues for social and b (%/year) 0.058a 0.100a 0.100a 0.058a 0.100a 0.100a
tm (year) 1965 1991 2002 1975 2004 2008

6 a
These results do not include the natural gas liquids. Fixed parameters.
150 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151

Table A.3
Parameters estimated for the multi-Hubbert model.

Scenario Parameters Coast Offshore shelf Selva

H1 H2 H3 H1 H2 H3 H1 H2 H3

a a a a a a a
EUR1 c (years) 86 21 86 50 21 50 21 13 50a
EUR (MBbl) 1046 91 345 215 144 214 449 470 909
Qm (MBbl/year) 15 5 5 5 9 5 27 45 23
b (%/year) 0.058a 0.238a 0.058a 0.100a 0.238a 0.100a 0.238a 0.383 0.100a
tm (year) 1955 1962 2011 1993 1977 2034 1996 1982 2030

EUR2 c (years) 86a 21a 86a 50a 21a 50a 21a 13 50a
EUR (MBbl) 1113 300 618 216 144 1662 456 472 2427
Qm (MBbl/year) 16 5 9 5 9 42 27 45 61
b (%/year) 0.058a 0.238a 0.058a 0.100a 0.238a 0.100a 0.238a 0.383 0.100a
tm (year) 1957 1962 2035 1993 1977 2055 1996 1982 2042

a
Fixed parameters.

References Gallagher, B., 2011. Peak oil analyzed with a logistic function and idealized Hubbert
curve. Energy Policy 39, 790–802.
Gestión, 2013. Perupetro confirmó que solo una empresa está interesada en lici-
Alejos, Ricardo, 2011. Documentos de Trabajo # 12: Proyecciones de la Matriz En- tación de lotes offshore. 〈http://gestion.pe/economia/perupetro-confirmo-que-
ergética al Largo Plazo. CEPLAN. Lima Peru. solo-empresa-esta-interesada-licitacion-lotes-offshore-2080369〉 (accessed on
Almeida, P., Silva, P.D., 2009. The peak of oil production—timings and market re- August 2014).
cognition. Energy Policy 37, 1267–1276. Gestión, 2014. Aprobar instrumentos ambientales en sector hidrocarburos debe
Amazon Watch, 2012. The Achuar and Talisman Energy. 〈http://amazonwatch.org/ demorar dos años, opina Perupetro. 〈http://gestion.pe/economia/aprobar-in
assets/files/2012-talisman-issue-brief.pdf〉 (accessed on March 2014). strumentos-ambientales-sector-hidrocarburos-demorar-dos-anos-opina-per
Anderson, K.B., Conder, J.A., 2011. Discussion of multicyclic Hubbert modeling as a upetro-2095309〉 (accessed on August 2014).
method for forecasting future petroleum production. Energy Fuels 25, González, E.M., Szklo, A., Lucena, A.F.P., 2013a. Uma aplicação do modelo multi-
1578–1584. Hubbert para a elaboração de cenários de produção de petróleo na Colômbia.
ANH, Agencia Nacional de Hidrocarburos. Cifras y Estadísticas: Producción 2012. Anais XV Congresso Brasileiro de Energia, vol. I. Segurança Energética e De-
〈http://www.anh.gov.co/ANH-en-Datos/Paginas/Cifras-y-Estad%C3%ADsticas. senvolvimento Econômico, Rio de Janeiro, pp. 73–85.
aspx〉 (accessed on December 2013). González, E.M., Szklo, A., Lucena, A.F.P., 2013b. Reservas, recursos e Últimos Re-
ANP, 2011. Plano Plurianual de Geologia & Geofísica da ANP 2007/2014. 〈http:// cursos Recuperáveis (URR): definições e aplicação no caso colombiano. Anais
www.anp.gov.br/?pg¼ 57732〉 (accessed on August 2014). XV Congresso Brasileiro de Energia, vol. I. Segurança Energética e Desenvolvi-
ARCH, Agencia de Regulación y Control Hidrocarburífero. Historial de Producción mento Econômico, Rio de Janeiro, pp. 55–72.
Nacional de Crudo Fiscalizado 2012. 〈http://www.arch.gob.ec/index.php/des Guseo, R., Dalla Valle, A., Guidolin, M., 2007. World oil depletion models: price
cargas/doc_details/479-historial-de-produccion-nacional-de-crudo-fiscalizado- effects compared with strategic or technological interventions. Technol. Fore-
2012.html〉 (accessed on December 2013). cast. Soc. Change 74, 452–469.
Bartlett, A., 2000. An analysis of US and world oil production patterns using Hub- Hamann, J., Paredes, C., 1991. Características estructurales, estrategias de creci-
bert-style curves. Math. Geol 32, 1–17. miento y tendencias de largo plazo de la economía peruana. In: Paredes, C.,
BCRP, 2012. Annual historic tables of the Peruvian economic activity. Central Re- Sachs, J. (Eds.), Estabilización y crecimiento en el Perú. GRADE, Lima,
serve Bank of Peru. 〈http://www.bcrp.gob.pe/estadisticas/cuadros-anuales-his pp. 55–107.
toricos.html〉 (accessed on December 2012). Haselip, J., 2011. Transparency, consultation and conflict: assessing the micro-level
Benes, J., Chauvet, M., Kamenik, O., Kumhof, M., Laxton, D., Mursula, Selody, J., 2012. risks surrounding the drive to develop Peru’s Amazonian oil and gas resources.
The Future of Oil: Geology versus Technology. IMF Working Paper. WP/12/109. Nat. Resour. Forum 35, 283–292.
Bentley, R.W., Mannan, S.A., Wheeler, S.J., 2007. Assessing the date of the global oil Hubbert, M.K., 1956. Nuclear energy and the fossil fuels. In: Meeting of the
peak: the need to use 2P reserves. Energy Policy 35, 6364–6382. Southern District, Division of Production, American Petroleum Institute. Shell
Blanchard, R., 2000. The Impact of Declining Major North Sea Oil Fields upon Future Development Company, San Antonio, Texas.
North Sea Production. 〈http://www.hubbertpeak.com/blanchard/〉 (accessed on Hubbert, M.K., 1959. Techniques of prediction with application to the petroleum
January, 2013). industry. In: Proceedings of the 44th Annual Meeting of the American Asso-
BP. Statistical Review of World Energy 2013. ciation of Petroleum Geologists. Shell Development Company, Dallas, TX, p. 43.
Brandt, A.R., 2007. Testing Hubbert. Energy Policy 35, 3074–3088. Hubbert, M.K., 1962. Energy Resources, A report to the Committee on Natural Re-
Brandt, A.R., 2010. Review of mathematical models of future oil supply: historical sources: National Academy of Sciences. vol. 54. National Research Council
overview and synthesizing critique. Energy 35, 3958–3974. Publication, 1000-D, Washington, DC, pp. 61–67.
Brecha, R.J., 2012. Logistic curves, extraction costs and effective peak oil. Energy IEA, 2013. World Energy Outlook 2013. International Energy Agency, Paris.
Policy 51, 586–597. INEI, Instituto Nacional de Estadística. 2014. Sistema de Información Económica. 〈http://
Calmet, F.N., 2012 Historia de la Industria de Petróleo en el Perú desde sus co- www.inei.gob.pe/estadisticas/indice-tematico/economia/〉 (accessed on August
mienzos hasta la fecha. 〈http://www.osinerg.gob.pe/newweb/uploads/GFH/His 2014).
toria%20del%20Petroleo%20Peru.pdf〉 (accessed on December 2012). Kaufmann, R.K., 1991. Oil production in the lower 48 states: reconciling curve fit-
Campbell, C.J., Laherrere, J.H., 1998. The end of cheap oil. Sci. Am. 278, 78–83. ting and econometric models. Resour. Energy 13, 111–127 ().
Campbell, C.J., 2013. Campbell’s Atlas of Oil and Gas Depletion, second ed. Springer, Laherrere, J.,1997. Multi-Hubbert modeling. 〈http://www.oilcrisis.com/laherrere/
New York, Heidelberg, Dordrecht, London. multihub.htms〉 (accessed on December 2012).
Campodónico, H.S., 1999. La inversión en el sector petrolero peruano en el período Laherrere, J., 2002. Is FSU oil growth sustainable? Pet. Rev., 29–31 (&35).
1993–2000. Serie Reformas Económicas, No. 23 (LC/L.1207), Santiago, Chile, Maggio, G., Cacciola, G., 2009. A variant of the Hubbert curve for world oil pro-
Economic Comission for Latin America and the Caribbean (ECLAC). duction forecasts. Energy Policy 37, 4761–4770.
CEPAL, 2013. Natural Resources: Status and Trends Towards A Regional Development Maple, 2012. Annual Report 2011. 〈http://www.maple-energy.com/downloads/
Agenda in Latin America and the Caribbean. 〈http://www.cepal.org/publicaciones/ MEP_%202011_Annual_Report_L.pdf〉 (accessed on August 2014).
xml/9/52079/CELAC-Naturalresources.pdf〉 (accessed on October 2014). Maple, 2013. Annual Report 2012. 〈http://www.maple-energy.com/downloads/
Chevalier, Jean-Marie, 2004. Les batailles de l'énergie. Folio. 2012_MEP_Annual_Report.pdf〉 (accessed on August 2014).
Defensoría del Pueblo, 2014. Reporte de Conflictos Sociales No. 122. 〈http://www. Maple, 2014. Annual Report 2013. 〈http://www.maple-energy.com/downloads/
defensoria.gob.pe/modules/Downloads/conflictos/2014/Reporte-Mensual-de- 2013_MEP_Annual_Report.pdf〉 (accessed on August 2014).
Conflictos-Sociales-N-122.pdf〉. Martínez, M.O., Napolitano, D.A., MacLennan, G.J., O’Callaghan, C., Ciborowski, S.,
El-Gamal, M.A., Jaffe, A.M., 2010. Oil, Dollars, Debt, and Crises: The Global Curse of Fabregas, X., 2007. Impacts of petroleum activities for the Achuar people of the
Black Gold. Cambridge University Press, New York, USA. Peruvian Amazon: summary of existing evidence and research gaps. Environ.
ESMAP, 1999. Joint UNDP/World Bank Energy Sector Management Program. Reform Res. Lett. 2, 045006.
and Privatization in the Hydrocarbon Sector, Peru. Maurer, N., 2011. .
Fattouh, B., 2007. The Drivers of Oil Prices: The Usefulness And Limitations of Non- MEF, 2011. Elaboración de la Nueva Matriz Energética Sostenible y Evaluación Ambiental
Structural Model, the demand–Supply Framework and Informal Approaches. Estratégica, como Instrumentos de Planificación.. Lima, Peru (in Spanish).
Oxford Institute for Energy Studies. WPM 32. MINEM, 2000. Statistical Yearbook Hydrocarbons 1999. Lima, Peru (in Spanish).
 M.F. Chavez-Rodriguez et al. / Energy Policy 77 (2015) 140–151 151

MINEM, 2005. Statistical Yearbook Hydrocarbons 2004, .. Lima, Peru (in Spanish). Reynolds, D.B., Kolodziej, M., 2007. Institutions and the supply of oil: a case study of
MINEM, 2011. National Energy Balance, 2010. Lima, Peru (in Spanish). Russia. Energy Policy 35, 939–949.
MINEM, 2013a. Statistical Yearbook Hydrocarbons 2012. Lima, Peru (In Spanish). Reynolds, D.B., Pippenger, M.K., 2010. OPEC and Venezuelan oil production: evi-
MINEM, 2013b. Yearbook of Hydrocarbons Reserves at December 31 of 2012, . in dence against a cartel hypothesis. Energy Policy 38, 6045–6055.
Spanish. Lima, Peru. Reynolds, D.B., Baek, J., 2012. Much ado about Hotelling: beware the ides of Hub-
MINEM, 2013c. Presentación del Ministro de Energía y Minas a la Comisión de bert. Energy Econ. 34, 162–170.
Energía y Minas del Congreso de la República. 〈http://www2.congreso.gob.pe/ Reynolds, D.B., 2014. World oil production trend: comparing Hubbert multi-cycle
sicr/comisiones/2013/com2013enemin.nsf//pubweb/4B510026CE75931005257 curves. Ecol. Econ. 98, 62–71.
BE4005113FC/$FILE/MEMPOLITICAENERGETICA.pdf〉 (accessed on June 2014). Sanchez-Sierra, G., Hughart, D., et al.,1984. Peru: Issues and Options in the Energy
Mohr, S.H., Evans, G., 2010. Long term prediction of unconventional oil production. Sector. Report No. 4677-PE. Report of the Joint UNDP/World Bank Energy Sector
Energy Policy 38, 265–276. Assessment Program.
Moroney, J., Berg, D., 1999. An integrated model of oil production. Energy J. 20, Saraiva, T.A., Szklo, A., Lucena, A.F.P., Chavez-Rodriguez, M.F., 2014. Forecasting
105–124. Brazil’s crude oil production using a multi-Hubbert model variant. Fuel 115,
Napolitano, D., Ryan, A., 2007. The dilemma of contact: voluntary isolation and the 24–31.
impacts of gas exploitation on health and rights in the Kugapakori Nahua Re- Schlumberger, 2007. Schlumberger, explorando las profundidades del Perú. Grafica
serve, Peruvian Amazon. Environ. Res. Lett. 2, 045005. Biblios, Lima, Peru.
Nashawi, I., Malallah, A., Al-Bisharah, M., 2010. Forecasting world crude oil pro- Smith, J.L., 2005. Petroleum prospect valuation and the option to drill again. Energy
duction using multicyclic Hubbert model. Energy Fuels 24, 1788–1800. J. 26 (4), 53–68.
Nystad, A.N., 1988. On the economics of improved oil recovery: the optimal re- Sorrell, S., Speirs, J., 2009. Review of Evidence for Global Oil Depletion. Technical
covery factor from oil and gas reservoirs. Energy J. 9, 49–61. Report 5: Methods of estimating ultimately recoverable resources. 〈www.ukerc.
Nystad, A.N., 1987. Rate sensitivity and the optimal choice of production capacity of ac.uk/support/tiki-download_file.php?fileId ¼ 289〉 (accessed on June 2014).
petroleum reservoirs. Energy Econ. 9, 37–45. SPE, 2007. Petroleum Resources Management System. 〈http://www.spe.org/in
Nystad, A.N., 1985. Petroleum taxes and optimal resource recovery. Energy Policy dustry/docs/Petroleum_Resources_Management_System_2007.pdf〉 (accessed
13, 381–401. on October 2014).
Orta-Martínez, M., Finer, M., 2010. Oil frontiers and indigenous resistance in the SPH, 2014. Libro Blanco de los Hidrocarburos. Sociedad Peruana de Hidrocarburos,
Peruvian Amazon. Ecol. Econ. Spec. Sect.: Ecol. Distrib. Confl. 70, 207–218. Lima, Peru. 〈http://www.sphidrocarburos.com/bnpscontenido.php?id_
Owen, N.A., Inderwildi, O.R., King, D.A., 2010. The status of conventional world oil cont ¼76〉.
reserves—hype or cause for concern? Energy Policy 38, 4743–4749. SUNAT, 2014. Estadísticas de Comercio Exterior. Anuario 2013. (accessed on August
Patzek, R., Croft, G., 2010. Global coal production forecast with multi-Hubbert cycle 2014).
analysis. Energy Policy 35, 3109–3122. Szklo, A., Machado, G., Schaeffer, R., 2007. Future oil production in Brazil—estimates
Perupetro, 2002. Maranon Basin: Technical Report. PARSEP: vol. I. Lima, Peru. based on a Hubbert model. Energy Policy 35, 2360–2367.
Perupetro, 2011. Promoting E&P in Peru’s mature and frontier basins. Profile: Per- Talara, Municipalidad Provincial de Talara, 2006. Compendio Histórico de Talara, p.
upetro. In: 20th World Petroleum Congress, Doha, Qatar. 25.
PERUPETRO, 2013. Public maps and drilling information delivered in digital format Tao, Z., Li, M., 2007. System dynamics model of Hubbert Peak for China's oil. Energy
by Perupetro. Policy 35, 2281–2286.
PERUPETRO, 2014a. Mapa de Lotes de Contratos, Cuencas Sedimentarias y Areas The Guardian, 2013. Indigenous Peruvians protest state oil company taking over
Naturales Protegidas. 〈http://www.perupetro.com.pe/wps/wcm/connect/ their land. 〈http://www.theguardian.com/environment/andes-to-the-amazon/
e9c77536-772e-4101-8f79-e5d21e2b966e/CNT þ Eneþ 2014.pdf?MOD 2013/jun/07/ peruvians-protest-oil-company-land-petroperu〉 (accessed on
¼AJPERES〉 (accesed on February 2014). March 2014).
PERUPETRO, 2014b. Estadística Petrolera Mensual 2014. 〈http://www.perupetro. Touzett, P. G., 2007. Marcos Regulatorios y el Rol de las Empresas Estatales de Hi-
com.pe/wps/wcm/connect/99663891-ee89-4cb6-ade5-681bdb4fa7cc/ drocarburos. Estudio de Caso: Peru. OLADE. 〈http://temp2.olade.org/doc
ESTADISTICAþ MENSUAL þMARZO þ 2014.pdf?MOD ¼AJPERES〉 (accesed on umentos2/hidrocarburos/MARCOS%20REGULATORIOS%20-%20ESTUDIO%20DE
May 2014). %20CASO%20PERU.pdf〉 (accessed on December 2012).
Petroperu, 2012. Hitos de la Historia (1968–2005). 〈http://www.petroperu.com.pe/ USGS, 2000. US World Geological Survey Petroleum Assessment – South America
portalweb/Main.asp?Seccion¼ 382〉 (accessed on December 2012). Province Boundaries, Washington, DC.
Petroperu, 2013. Instalaciones: Oleoducto Norperuano. 〈http://www.petroperu. USGS, 2012. An estimate of undiscovered conventional oil and gas resources of the
com.pe/portalweb/Main.asp?Seccion ¼76〉 (accessed on July 2013). world, 2012: U.S. Geological Survey Fact Sheet 2012–3042, 6 p. 〈http://pubs.
Pontoni, A. 1982. Políticas Petroleras en el Perú: 1968–1982. Documento de Trabajo usgs.gov/fs/2012/3042/〉 (accessed on October 2014).
No 53. 〈http://departamento.pucp.edu.pe/economia/images/documentos/ USGS, 2013. Technically Recoverable Shale Oil and Shale Gas Resources: An As-
DDD53.pdf〉 (accessed on December 2012). sessment of 137 Shale Formations in 41 Countries Outside the United States.
Quijandria, J., 1993. El futuro de la empresa del Estado: el caso del sector petrolero. Washington, DC.
IEP, Documento de Trabajo, 49. Serie Documentos de Política, 3. Lima, Peru. Vásquez, A.C., 2005. La Organización Económica de la Industria de Hidrocarburos
Ramcharran, H., 2002. Oil production responses to price changes: an empirical en el Perú: el Segmento Upstream del Sector Petrolero. Working Paper No. 8,
application of the competitive model to OPEC and non-OPEC countries. Energy Office of Economic Research, OSINERGMIN.
Econ. 24, 97–106. Wang, J., Feng, L., Snowden, S., Wang, X., 2011. A comparison of two typical multi
Reátegui-Zirena, E., Stewart, P., Whatley, A., Chu-Koo, F., 2012. Acute Toxicity and cyclic models used to forecast the world’s conventional oil production. Energy
Mutagenicity of Peruvian Crude Oil and Oil-Contaminated Samples from the Policy 39, 7616–7621.
Peruvian Amazon. Instituto de Investigaciones de la Amazonía Peruana. Watkins, G.C., 1992. The Hotelling principle: Autobahn or Cul de Sac? Energy J. 13,
Rehrl, T., Friedrich, R., 2006. Modeling long-term oil price and extraction with a 1–24.
Hubbert approach: the LOPEX model. Energy Policy 34, 2413–2428.
Reynolds, D.B., Zhao, Y., 2007. The Hubbert curve and institutional changes: how
regulations in Alaska created a US multi-cycle Hubbert curve. J. Energy Dev. 32,
159–186.

View publication stats