Monday, March 13, 2006

Mexico's Ability to Export Oil

This post is the third in the same topic, previous posts were the following:

Saudi Arabia
Iran

Mexico is the second largest source of crude imports for the US with 1.55 mbpd in 2005 [5]. I tried to compile different URR (Ultimate Recoverable Ressource) estimates from various sources:

SourceReservesYearCumulativeYTFUltimate
BP Statistical Review14.803
Year-End 200433.9
5.975
54.67
Oil & Gas Journal12.882
January 1, 200635.8
5.975
54.67
World Oil14.803
Year-End 200433.9
5.975
54.67
IHS 30
Year-End 200227
14
71-99
USGS25.1
Year-End 199623.6
18.60354.7-88.9
ASPO28.6
200223
4.250
volumes are in Giga-barrels (Gb)

The production profile is taken from BP Statistical Review of World Energy (june 2005) and completed for years prior to 1965 by using the ASPO graph shown on Figure 1. (Note: I'm not sure that this production profile includes all liquids).

Fig 1. Oil production of Mexico (src: ASPO)

From the table above, we can design a simple triangular probability density function for the URR where the maximum of probability corresponds to the median value.

Fig 2. URR probability distribution function derived from the table above.

In order to model future production, I applied the SBM-PF method and the Hubbert Linearization technique. The results are shown on Figure 3 and 4. Both methods give similar result for the future production profile which is predicted to decline from 2006 with a decline rate between 7% and 8%.

Fig 3. Future production based on the SBM-PF method and the Hubbert Linearization technique (see Figure 4). Both methods give an URR around 72 Gb. Production is peaking in 2005-2006.

Fig 4. Same as Figure 3 but using a P/Q versus Q representation.

The domestic oil consumption has grown rapidly but is showing signs of a slow down since 2000 (see Figure 5). Note also that retail prices for Diesel and super-gasoline have increased since 2000 (see Figure 6).

Fig 5. Mexico's oil consumption (millions of barrels per day, src: BP)

Fig 6. Green Benchmark Line = Retail Fuel Prices in the UNITED STATES. Red Benchmark Line = CRUDE OIL Prices on World Market ("Brent" at Rotterdam)[3]

The annual population growth rate is about 1.9% [1]. The population was 107.0 million in mid-2005 and is expected to reach 129.4 million in 2025 (. If we divide the oil consumption by the population numbers, we can see that the oil consumption seems to have reach a ceiling since the 90s at about 6.5 barrel/capita/year (Figure 8).

Fig 7. Mexico population trend (src: UN)

Fig 8. Number of barrels consumed per capita per year.

Based on a constant consumption per capita (blue dotted line on Figure 8) we predict Mexico's future oil consumption based on the UN population forecast shown on Figure 7. The result is shown on Figure 9 for the three population model variants. The oil production forecast for Mexico was derived from the SBM-PF model. We can observe that the oil exports are expected to go down by 2006-2007. The fraction of oil exports may fall below 40% after 2015 (Figure 10) and the US could lose an important and reliable source of oil imports (92% of Mexico's exports are for the US).

Fig 9. Projected oil consumption based on the population growth and a constant consumption per capita per year (6.5 barrels/capita/year).

Fig 10. Projected share of the oil exports for the different population scenario. The export share will reach a maximum between 2006 and 2007.

References:

[3] www.internationalfuelprices.com

Monday, March 06, 2006

M. King Hubbert's Lower 48 Prediction Revisited

What can 1970 and Earlier Lower 48 Oil Production Data
Tell Us About Post-1970 Lower 48 Oil Production?

Jeffrey J. Brown & "Khebab"
March 6, 2006

Introduction

Fifty years ago this week, on March 8, 1956, at a meeting of the American Petroleum Institute in San Antonio, Texas, M. King Hubbert, in the preprinted version of his prepared remarks, had the following statement, "According to the best currently available information, the production of petroleum and natural gas on a world scale will probably pass its climax within the order of a half century (i.e., by 2006), while for both the United States and for Texas, the peaks of production may be expected to occur with the next 10 or 15 years (i.e., 1966 to 1971)." As more and more people are learning, Lower 48 oil production, as predicted by Dr. Hubbert, peaked in 1970, and it has fallen fairly steadily since 1970.

Kenneth Deffeyes, in Chapter Three of his recent book, "Beyond Oil: The View From Hubbert's Peak," described a simplified way of predicting the production peaks for various regions and for their subsequent declines. One simply plots annual production (P) divided by cumulative production to date (Q) on the vertical axis, or P/Q, versus Q on the horizontal axis (see Figure 1). Stuart Staniford, on The Oil Drum Blog, has described this technique as "Hubbert Linearization" or HL [1].

With time, a HL data set starts to show a linear progression, and one can extrapolate the data down to where P is effectively zero, which gives one Qt, or ultimate recoverable reserves for the region. Based on the assumption that production tends to peak at about 50% of Qt, one can generate a predicted production profile for the region. The Lower 48 peaked at 48.5% of Qt.

Using the HL technique, Dr. Deffeyes, an associate of Dr. Hubbert, predicted that the world crossed the mathematical 50% of Qt mark on December 16, 2005. In other words, Dr. Deffeyes believes that the world is now where the Lower 48 was at in the early Seventies.

Given this prediction, we wondered how accurate has the HL technique been in predicting post-peak cumulative Lower 48 oil production. Also, what can the Lower 48 model tell us about the future of world net oil export capacity? Finally, we have a recommendation, a fundamental shift in US tax policy.

Hubbert Linearization Analysis of Post-Peak Lower 48 Oil Production

We generated a HL plot using annual data from 1942 through 1970 using data from the EIA [2][3]. This plot predicted a Qt of approximately 196.1 Gb (Figure 1). We then converted the data to the typical production versus time plot (Figure 2).

Fig 1. Hubbert Linearization technique applied to the Lower-48. Only the data between 1942 and 1970 are used to perform the fit (green points).

Fig 2. Lower-48 production data and logistic curve resulting from the Hubbert Linearization technique.

Using only 1970 and earlier production data, the HL plot predicted that post-1970 Lower 48 cumulative oil production would be 78.5 Gb, which is 98.7% of the actual post-1970 cumulative oil production (79.5 Gb). Note that actual oil cumulative production was slightly less than what the HL plot predicted.

This suggests that we can almost certainly place a good deal of confidence in Dr. Deffeyes' prediction that the world is on the edge of a permanent and irreversible decline in conventional oil production. If the Lower 48 were the sole source of crude oil for the world, for every two gallons of gasoline that the world bought in 1970, we would be bidding for one gallon today.

World Net Oil Export Capacity

According to the EIA, in 2004 the top four net oil exporters were Saudi Arabia; Russia; Norway and Iran, collectively accounting for more than half of the net exports from the top net oil exporters (those with net exports of one mbpd or more).

Using all available data, we generated a HL plot for the sum of the production from these four countries, deriving a Qt of 569.8 Gb (Figure 3). We then generated a production versus time plot (Figure 4), showing the historical and predicted future production for all four countries. Note that based on the HL analysis, these four countries are collectively about 55.5% depleted. Keep in mind that the Lower 48 analysis was 98.7% accurate in predicting post-1970 cumulative Lower 48 production (using only data through 1970).

Fig 3. Hubbert Linearization technique applied to the total production of the top four oil exporters. Only the data between 1979 and 2004 are used to perform the fit (green points).

Fig 4. Sum of the top four production data and logistic curve resulting from the Hubbert Linearization technique.

The predicted Russian production profile is the most controversial of the four plots that we analyzed. However, Russian Minister of Industry and Energy Viktor Khristenko, in an interview with Russian Profile [5], had the following statement, "One important point is that the longer we delay making this decision (encouraging frontier exploration in Russia with tax breaks), the harder it will be to feel the effectiveness of the measures taken: the structure of the country's reserves will continue to get worse and Russia could end up facing a real collapse in oil production."

A critical point to keep in mind is that an exporter can only export what is left after domestic consumption is satisfied. Consider a simple example, a country producing 2.0 mbpd, consuming 1.0 mbpd and therefore exporting 1.0 mbpd. Let's assume a 25% drop in production over a six year period (which we have seen in the North Sea, which by the way peaked at 52% of Qt) and let's assume a 10% increase in domestic consumption. Production would be 1.5 mbpd. Consumption would be 1.1 mbpd. Net exports would be production (1.5 mbpd) less consumption (1.1 mbpd) = 0.4 mbpd. Therefore, because of a 25% drop in production and because of a 10% increase in domestic consumption, net oil exports from our hypothetical net exporter dropped by 60%, from 1.0 mbpd to 0.4 mbpd, over a six year period.

We are deeply concerned that the world is probably facing an imminent and catastrophic collapse in net oil export capacity because of declining production and increasing domestic consumption in the top exporting countries. Figure 4 predicts that the combined oil production by the current top four net exporters will be reduced by 50% by the year 2028, which would probably equate to a reduction in net oil exports of at least 75% over the next two decades.

As the reality of declining conventional oil production sets in, enormous amounts of capital and energy will be allocated toward nonconventional oil reserves in an effort to try to keep the supply of liquid transportation fuels--gasoline; diesel and jet fuel--growing. The principal effort will be focused on very heavy oil, bitumen (tar sands) and coal liquefaction. These nonconventional sources will help, but the key point to keep in mind is that we are merely discussing increasing the rate of extraction of fossil fuels. The question we have to ask is whether the enormous amounts of capital and energy might be better spent on renewable energy projects. In any case, we predict that nonconventional sources of oil will only serve to slow--and not reverse--the long term decline in world oil production.

Proposal: Replace the Payroll Tax with an Energy Tax

Some have argued that the suburbs are dead; the suburbanites just don't know it yet. It's probably more accurate to say that the suburban commutes are dead; the suburban commuters just don't know it yet.

We recommend that the United States abolish the payroll tax (Social Security + Medicare tax) and replace it with either a liquid transportation (petroleum) fuel tax or an overall (nonrenewable) energy tax.

The majority of American households pay more in the payroll tax than in the income tax. This would be a tax cut for most households and it would a massive tax increase on those who are profligate in their use of energy. No matter where one lives, the cost of goods would go up, but if you lived close to where you work, your effective tax rate would go down. Of course, those who persisted in long commutes would pay the price.

There would of course be very powerful forces opposed to this idea--the housing industry; auto industry; airlines; trucking--the list goes on. But the fates of these industries are sealed. It's not a question of if they will contract; it's just a question of when. The sooner it happens, and the sooner these industries start emphasizing energy efficiency, the better off we all will be.

A high gasoline tax does not necessarily equate to a lower standard of living. Norway, with the highest gasoline tax in the world, has the highest standard of living in the world, perhaps partly because their car ownership per 1,000 people is about half of what it is in the US.

There would be some other benefits. As we turned to walking, biking and mass transit, our health would improve. There is pretty much a linear correlation between obesity rates and total miles driven (here in the US, we are the world champs in both categories). In addition, since this is in effect a consumption tax, everyone who now avoids paying Social Security taxes would no longer be able to avoid paying them.

However, the primary reason for implementing the proposal is that it would cause an immediate and massive across the board push for greater energy efficiency and it would unleash enormous free market forces against profligate energy use.

Jeffrey J. Brown is an independent petroleum geologist in Addison, Texas
E-mail: westexas@aol.com

Khebab has a Masters Degree in Physics and a Ph.D. in signal processing. He is currently a researcher in a computer vision lab in Quebec, Canada, with ten years experience in R&D.

References:

[5] russiaprofile.org: The Need for Energy Dialogue

Related Posts:

Saudi Arabia's Ability to Export Oil
Iran's Ability to Export Oil
Uncertainties About Russian Reserves and Future Production

Saturday, March 04, 2006

How Probable is a US Recession if the Oil Supply is Shrinking?

This post is a little addition to the previous post "US Addiction to Oil". We looked at the annual variations of US oil consumption per capita and observed a significant correlation with the annual GDP per capita variations (see Figure 1 below). I propose to investigate a little bit more into that relationship. We note BCYV the number of Barrels per Capita per Year Variation and GCYV the GDP per Capita per Year Variation. Valleys on the GCYV curve are economic slowdown and recessions that have occured many times in the past when GCYV becomes negative. I surimposed on the graph the different major recession occurences that have affected the US economy. There was only two major oil supply contractions in the past: the 1973 oil embargo and the 1979-1980 Iranian crisis. They correspond to the two big drops below -4% on the BCYV curve.

Fig 1. Year-on-year variations of the GDP per capita (in blue) and the consumed oil barrels per capita (in red). The light dotted gray lines indicate recessions.

We use a bi-dimensional gaussian function to describe the joint probability distribution of GCVY and BCYV as shown on Figure 2.

Fig 2. Scatter plot of GCYV vs BCYV and the multivariate gaussian probability function. The blue dotted line indicate the principal direction of variation (first eigenvector).

The parameters (covariance matrix and mean vector) for the bi-dimensional gaussian estimated from the data are the following:

The blue line on Figure 2 is the principal component of the covariance matrix and corresponds to the following equation:

So, in average the US have enjoyed a nice 2.15% GDP growth since 1965. Now, let's compute some probability values from that model such as the probability of a recession (we note B=BCYV and G=GCYV):

We repeat this calculation for the case where BCYV > 0, the different values are shown on Figure 3. If the oil consumption per capita is shrinking we have a 30% chance of having a recession whereas the probability is only 2% if the oil consumption per capita is growing.

Fig 3. Probability values of a recession or a GDP growth having different oil consumption variations.

Of course, this a simplistic analysis and there are many other factors that can influence the GDP growth. Also, we didn't take into account the impact of energy conservation, efficiency and energy alternatives.

References:

Wednesday, March 01, 2006

Hydro-Quebec Will Increase its Electricity Rates by 5.3%

Hydro-Quebec said yesterday it will increase its electricity rates by 5.3 per cent - nearly double the national inflation rate of 2.8 per cent - effective April 1.

src: The Montreal Gazette

The reason for the increase is given on the Quebec Energy Agency website:

For 2006, the Distributor’s revenue requirement is $10,051 million, necessitating additional revenues of$463 million and therefore an average rate increase of 5.3% for the 2006-2007 rate year.
The increase is to cover its capital and operating costs that have reached 11 cents/kWh because electricity demand has soared in Quebec.