Voltaire, Patron Saint of Bloggers

“…And this public, he knew how to get hold of it; a public intelligent and fickle, inquisitive and sophisticated, whom one trifle would displease and another trifle amuse, a public with a narrow and delicate taste, with a short attention span, which one must constantly catch and intrigue. Every single day for twenty-three years he served up to it the sauce of wit, satire, jokes, and smut with which it was necessary to season his ideas.

Above all, he wrote clear, short, and quick. No more big works. Little twelve-page tracts, leaflets a couple pf pages long. “Twenty volumes in folio…will never make a revolution; it’s the little pocket-volumes at thirty cents apiece that have to be watched. If the New Testament had cost 4,200 sesterces, the Christian religion would never have taken root.” These “little pot-pies,” these portable scandal sheets, easy to read, and continuously exciting,…emerged in all forms, on all subjects, in verse, in prose, dictionaries, stories, tragedies, diatribes, extracts on history, literature, metaphysics, religion, the sciences, politics, legislation, Moses, snails, Shakespeare, and notes written by a gentleman. In reality, dearly as he prized the arts of literature and poetry, they became nothing more for him than a means to an end. Tragedies and verses served to hasten the spread of his ideas.

He repeated himself, he went over the same ground again and again. He was aware of it, and started the same ideas on still another round. For he knew that ideas enter the public mind only by dint of repetition. But the seasoning must be varied, to prevent disgust; and at that art he was a past master.

He as all the qualities, with many of the faults, of the journalist, above all the gift for the immediate, and the penetrating voice which carries and fixes our attention through the noisy confusion of life. But it is not enough to say Voltaire is a journalist; all by himself he is a journal, a great journal. He does the whole thing himself, the serious articles, the spot-reporting, the gossip column, the funny papers, the crossword-puzzles. He is a journal, but also a review, an encyclopedia; all the jobs of popularisation, propaganda, polemic, and information fall together in his hands. This quick old man is a whole press, a complete popular library.

Finally, by means of his innumerable letters, which reached people of every rank and every nation -the king of Prussia, the Empress Catherine, German princes, Russian or Italian gentlemen, English thinkers, ministers, courtiers, provincials, judges, comedians, abbés, men of letters, administrators,  merchants, lawyers, women of the world -by these thousands of letters one which does not contain a compliment to the addressee’s self esteem, a joke for his amusement, and a thought for him to mull over, Voltaire interested I know not how many individuals in the success of his propaganda. He made them carriers, voluntary and uncontrollable, of his ideas. He strengthened, he doubled, by means of his correspondence, the effect of his pamphlets.”

-Gustave Lanson, Voltaire at Les Délices and at Ferney.

 

 

How is oil discovered?

Feasibility Overview

 The feasability assessment I’m attempting works like this:

1. Define a likely range for the rate of oil decline (in Giagjoules/person/year)
2. Define a likely range for the rate of substitution for potential technology, individually and in aggregate (also in Gigajoules/person/year)

There are 3 possible outcomes:

A. 1 is greater than 2, in which case peak oil can be considered a problem. We can then move discussion to the impacts of peak oil on the economy.

B. 2 is greater than 1, in which case peak oil is not so much of a problem.

C. Something went wrong in the analysis.

The next couple of posts in the feasibility assessment series will look at the likely timing and decline rate of conventional crude oil in order to define (1) above. To do that we need to know the following:

1. How is oil discovered? Gotta find it first.
2. How is oil extracted? Yep.
3. How does reserve growth work? A tricky issue that needs its own post.
4. How much oil is there?

Discovery

This post takes a look at oil discovery. I’m going to crib off WebHubbleTelescope (Web, for short) over at mobjectivist (sidebar) whose ‘Shock’ model takes what we know about the 4 points above and applies math. Web points out that the maths isn’t particularly hard, but he probably means ‘not that hard, if you studied maths at university level’,  so I’ll break it down with analogies (for my own understanding more than anything else). Below is a chart showing global backdated discovery vs. oil extraction. Notable for the fact that we’re just not discovering like we used to, and we’ve been extracting more than we’ve been discovering:

Readers of mobjectivist will know that an oil discovery model that describes the data can be built from just two moving parts:

1. an accelerating and dispersed search for oil.
2. dispersed oil fields.

The Search for Oil

1. Why should the search for oil accelerate? Answer: Maximum Power.

Oil is highly useful. It’s liquid which makes it easy to transport. It’s energy dense, which makes it good for powering moving vehicles. When mixed with air it’s explosive which makes it good for releasing a lot of energy very quickly. But perhaps most importantly at least  in the early days of oil discovery it had a very high Energy Return on Investment ratio (EROI). For every 1 barrel of oil worth of energy invested into prospecting, and pumping oil, oil companies in the 1930′s could expect to extract 100 barrels of oil ♣. High EROI resources are more likely to be scaled up quickly and gives industrial society the energy equivalent of a sugar rush.  Oil extraction delivers much more available energy to society. A small fraction of that can be devoted to acquiring even more energy, accelerating the search. Humans in competitive environments will use energy consumption as their main status proxy. Nate Hagens has blogged extensively on this here. Ayres and Van de Bergh (2005)♥ describe how energy supply and energy demand are in a non-equilibrium feedback loop that drives economic growth (Figure 1. Inside loop). I quote them extensively below:

Figure 1. The feedback loops of economic growth

“This resource-driven feedback mechanism for growth is indicated by loop 1 in Figure 1. It can be described briefly as follows: technological progress has made fossil fuels steadily and dramatically cheaper and more convenient to use since the early 18th century. This, in turn, encouraged the substitution of fossil fuel-derived energy and mechanical power for work by animals and humans…Both cheaper fuels and better metals made it possible to construct better, cheaper, and more efficient machines, including steam engines and machine tools. This, in turn, permitted continuous and drastic further reductions in the cost of mining and transporting coal (later other fuels), and the delivery of mechanical power to users, including the coal mines and the transport systems themselves…

…Conceptually, the cycle consists of two separate elements. First, economic growth since 1800 has been driven, to a large extent, by utilizing machines (steam engines, internal combustion engines, and tractors) powered by fossil fuels as a substitute for, and multiplier of, human and animal labor. Second, the extensive use of fossil fuel-derived chemical fertilizers and pesticides on farms is another, more recent, technique of increasing productivity by using less labor. Naturally, as resource extraction and conversion costs fall due to economies of scale and learning-by-doing, economic growth is stimulated, resulting in a further increase in the overall use of raw materials and fossil fuels (this is the so-called ‘rebound effect’ writ large). In other words, a positive feedback mechanism is operative…

…It is important to emphasize that this feedback cycle is not merely a particular form of learning-by-doing, nor is it fundamentally attributable to scale economies, although both learning and scale are obviously involved and can reinforce it. One of the two key elements of the cycle is the availability, at ever-lower costs, of fossil fuels, initially coal, and subsequently petroleum and natural gas or nuclear energy. These are, of course, material resources. But they differ from other resources, such as construction materials, in that they are not embodied in products (except for plastics and synthetic fibers). They are entirely consumed for the purpose of generating heat, mechanical power, or electric power.”

A positive feedback loop between energy demand and supply is the essence of Maximum Power and stimulates an accelerating search for oil. As more people find oil useful, more knowledge is accumulated on how to find it, and more people get into the oil game. In some countries the search proceeds quickly (USA), while others take their time (Iraq), leading to varying rates of search.

Dispersion

 2. What does dispersion mean and why are oil fields dispersed?

Take a look at the satellite photo below. It shows a lake prone region of Ontario, Canada. What do you see?

If you answered “lots of little lakes, a couple of mid-size lakes and 1 big lake (Trout lake) you’re on the right track. Lake size distribution is dispersed and you find the same thing with oil fields. Most oil comes from a few supergiant fields but most fields are much smaller. Web has plotted field sizes vs. field rank in the logarithmic chart below. The vertical axis is in logarithmic million barrels. From the chart we can see that the top 10 oil fields in the world are all over 20 Billion barrels in size. The next 90 largest fields are between about 20 billion and 2 billion barrels etc.

Combining the 2…

So when you have an accelerating searches (of variable rates), for fields of dispersed size, the discovery rate over time will look something the orange curve below:

I’ve taken the curve that Web describes here  as an integration of all of the above and fitted it to backdated discovery data from here [xls]. Backdated discovery is really two processes (the initial discovery plus subsequent reserve growth backdated to the date of the initial discovery) so we can’t use this model for predicting future discovery. But we can see how this parsimonious model fits the data quite well. There’s a lot of noise, but that’s as we’d expect with such a large variation in field size.

An alternative?…

Its hard to find any correlation between price and discovery rates, so arguments that we will discover enough oil to prevent peaking as the price goes up are not reflected in the data (NB: to significantly delay peaking we would need to discover more than we consume, something that hasn’t consistently occurred for 30 years). The alternative explanation for the recent falling discovery is that OPEC and nationalised oil companies restricted discovery from the 1970′s onwards. In this narrative, without the discipline of competing in a free market nationalised oil companies have complacently allowed their exploration programs to go to pot. If only these countries would open up to the liberation of the free market, then the glory days of oil discovery will begin anew! I’ll chew over that one over the summer break, and get back to you  in the new year…

♣ Hall, Powers and Schoenberg (2008) Peak Oil, EROI, Investments and the Economy in an Uncertain Future. Biofuels, Solar and Wind as Renewable Energy Systems

♥ Ayres and Van de Bergh (2005) A theory of economic growth with material/energy resources and dematerialization: Interaction of three growth mechanisms. Ecological Economics. Vol 55, Iss 1. http://dx.doi.org/10.1016/j.ecolecon.2004.07.023

Which is hurting more, oil prices or debt deleveraging?

Regular programming has been interrupted this week by a surge in interest in macroeconomic theory and peak oil in the NZ blogosphere! Mostly this interest has been catalysed by last Friday’s sustainable economics conference*, which I had hoped to attend. Unfortunately, I had a work thing, and missed it, but my spies reported it went reasonably well.

Attendees Matt Nolan* and  Claire Browning* covered the conference . Claire has built on a earlier post* on the prospects of Solid Energy converting Southland lignite to liquid fuel. Over at The Standard, shadowy pseudonymous blogger ‘Marty G’ posted* Jeff Rubin’s ASPO conference speech transcript on oil prices as the major factor of the financial crisis. Followed soon after by a thoughtful piece* (and series beginner) on what is the economy for anyway? Meanwhile, buried in the Monday lunchtime tweetstream, iconoclast financial journalist Bernard Hickey crossed over* into full peaksterdom (/peaknikhood?/ peakerstroika?).

The surge in interest in topics dear to my heart is most welcome. Maximum Power usually follows my programme of wonkish arguments, but this recent interest has spurred me to enter the fray of inter-blog discussion. I offer two points:

First, the question I would have asked conference panelists. I can’t speak about this one, but in my experience conferences of this sort typically focus on how to transition our economy wholesale into ‘sustainability’, which is usually abstractly defined. While the transition is thought of as a journey of a thousand miles, perhaps beginning with a few small steps, except that the road turns out to be pea-soup foggy and nobody really has any directions. My question is supposed to call attention to the scree slope that we seem to have stumbled down:

“Given that we seem to be in the midst of a 1 in 80 year debt deleveraging cycle and that the world’s major energy resource is about to go into terminal decline, what is the risk that New Zealand’s GDP/cap will never recover to 2008 levels? 1%, 5%, 50-50?”

Probably wouldn’t have gone down very well. A question like this however, does focus the mind on the here and now, and an emergency response approach to policy follows from it, rather than the usual bland commitment to the environment.

Second, I disagree with Jeff Rubin’s analysis of the Global Financial Crisis (GFC). Sort of (I disagree with conventional wisdom too.) Rubin is sceptical of subprime mortgages as the ultimate cause of the GFC and instead pins the blame on the rise in oil prices 2002 through 2008. I agree that the subprime mortgages are a sideshow, but I even though I’m blogging on peak oil I think Rubin’s case is too simplistic.

The clearest and most cogent explanation of the GFC I have found  is Prof Steve Keen’s (see sidebar). Keen does his economics in a Post-Keynesian framework which has the bulk of money created ‘endogenously’ by private bank lending rather than the fiat money creation of central banks that most economists learn in their textbooks. In this framework, the money supply expands and contracts on borrower sentiment. When people are (over)confident about the future they have no problem taking out a loan, to say, invest in the housing market because house prices always go up, right? Debt piles up, until something bad happens and borrowers get a bit nervous about how much they owe, how much interest they’re paying, and how much they could be paying if interest rates rise. They decide to reverse their previous position! They begin to pay down debt, diverting money that would normally feed consumption to saving.  When borrowers all  do this together it worsens the economic situation and leads to further uncertainty and further declines in consumption.

(The theorists behind this framework include, Irving Fisher, Hyman Minsky, Basil Moore, Joseph Schumpeter, Keen himself, and some chap called Karl Marx if you want to look them up). Most orthodox economists are not comfortable with this framework, because it breaks a several orthodox doctrines like fiat money creation, rational expectations, and economic equilibrium (which I’ll deal with later).

To track this debt contraction/expansion dynamic Keen measures the total amount of Debt compared to GDP. I’m going to show how NZ figures measure up, but the reader should keep in mind that most advanced economies have similar levels of debt. The following chart shows these figures from the RBNZ for the New Zealand Economy.

Business debt has been deleveraging since the end of 2008, while other private sectors are only just starting to turn in the most recent quarterly figures (July, 2010). Government debt has been expanding at a fair clip but is still lower than recent historical levels. The standout is of course, mortgage debt which has rocketed up to 90% of GDP with only a brief pause at halftime.

Keen also uses an ‘Aggregate Demand’ metric. Aggregate demand is GDP + the change in debt and is supposed to measure demand for goods and services (included in GDP) and assets like houses and equities (not included in GDP). Aggregate Demand is a bit tricky and is used more as a ‘roughly right’ rather than a ‘precisely wrong’ measure. Keen discusses the ins and outs here. The key point is that when debt growth becomes large relative the economy, any slowdown of debt accumulation reflects a change in sentiment that flows through to less demand for goods and services. The New Zealand economy continued to accumulate debt through the financial crisis, but at a slower rate than the early Noughts. Here’s the chart for Aggregate Demand (click to embiggen):

Change in debt figures in this chart are quarter over same quarter year previous year, so the chart still shows a growth in debt (the red line is above the blue line). You can see that aggregate demand has fallen since the start of 2008, propped up by government borrowing (the gap between the red and green lines). When aggregate demand falls and the economy has become dependent on debt-financed spending, then some people lose their jobs. The next chart shows the correlation between debt-financed proportion of aggregate demand and unemployment (inverted):

So how does this relate to oil prices? On the one hand oil expenditures are much smaller than debt-financed expenditures. The next chart shows both in relation to Aggregate Demand:

Oil expenditures come from infoshare’s import data and are summed crude oil and petroleum products figures. I’ve drawn a line from the respective high and low points and use the difference to create the next chart:

The rise in oil prices is not insignificant. At the price peak in 2008 we were sending about 8 billion dollars a year overseas to pay for our oil imports, which is about the size of our dairy exports. We also know that when oil prices go up they often have a larger than expected impact on spending as consumers delay purchases on big-ticket items. However, the slowdown in debt growth (and recent beginnings of a debt contraction) has been substantially greater. What I’m suggesting is that while oil prices have certainly contributed to economic uncertainty,  and uncertainty bursts overconfident debt bubble, overconfident bubbles have to burst sometime.

Clearing this debt overhang is going to take a long time. And of course, if anything bad happens in the next 2-5 years, like say, oil extraction rates coming off their current plateau, then we’re in completely different territory.

Oil Markets I

In the Oil Markets series of posts I look at the problems of the Efficient Market Hypothesis (EMH), and present an alternative, the Adaptive Market Hypothesis.

The Efficient Market Hypothesis

To recap, the EMH is a mechanism that justifies the view that the rate of substitution will at least match the rate of oil decline. In the efficient market view, markets will signal the impending scarcity of oil by smoothly raising the price, allowing investors sufficient time to invest in alternatives. The EMH states that the price of an asset reflects all available information about that asset. Richard Thaler splits the EMH into two parts:

1. ‘There is no free lunch’ Because the price reflects available information, over the long-run you cannot beat the market. You’re better off investing in a diversified portfolio, than trying to second guess the market and predict where the price will go.

 2. ‘The price is right.’ Because the price reflects available information, markets are allocatetively efficient -that is, we should allow the oil market, rather than the government to allocate investment in energy supplies and energy consuming capital (cars, roads etc.). Except for roads and parking, this is current government policy, and the policy under the previous Labour government (A discussion of schizophrenic government policy will have to wait for another time).

Applying the EMH to the oil market a cornucopian might tell the following story:  ’as information about oil scarcity becomes available, it informs the position of a trader who buys and sells oil futures contracts. Futures contracts are in effect promises to purchase a certain amount of oil, at a certain price, on a certain future date. Futures allow businesses that use a lot of oil, such as airlines, to ‘hedge’ against a large increase in oil prices. An oil price spike makes running an airline difficult and hedging smooths out the price, allowing the airline to follow a business plan, but financial traders also participate in the oil futures market.  The oil futures market and regular oil market move in tandem. A trader, convinced that peak oil is a reality and that the price of oil will be more expensive in the future, realises that oil futures contracts are undervalued and buys them with the intent to sell when the price is higher (arbitrage). As this strategy proves profitable, other traders quickly adopt it. Willing buyers of cheap futures increase while willing sellers decrease. Thus the market clears at a higher price, and effectively aggregates available information about oil scarcity. This allows the price to rise before oil scarcity bites, smoothing out price increases and signalling profit opportunities to intrepid investors in alternatives to oil. Because the futures price out to 2018 ($92/barrel) is currently only slightly  higher than the spot market ($86/barrel), we can surmise that most of these well informed, professional traders do not agree that oil is about to get severely scarcer. Who are you to say differently?’

We’ll call the story above the efficient market view. Versions of the efficent market view are common throughout the blogosphere, and indeed, official government publications*. Often the discussant suggests that peakniks cannot be sincere in their beliefs because they are not using their special knowledge of the future of oil to turn a profit through purchasing oil futures cheaply.

I will not attempt to prove the EMH wrong. Instead, I will attempt to sketch out an alternative explanation of oil trading. Opponents of the EMH (and there are plenty, even in orthodox circles) like to point to dramatic bubble burstings such as the dotcom or housing  busts, as episodes that prove the EMH wrong. In reply, an EMH proponent can argue that these episodes new information about housing or internet startups was revealed, and the market ‘self-corrected’. This is an airtight argument, although in my mind reminiscent of young earth creationists arguing that God put those fossils there to test their faith.  

Scarcity

The problem with the efficient market view is that the knowledge of a likely decline in oil extraction rates does not translate into certainty of the oil price levels at a particular time intervals, and any extraordinary arbitrage opportunity. Peakniks propose that oil will get scarcer. This has usually been interpreted to mean that the price of oil will be higher. While I generally expect higher prices, I also expect lower wages from the economic impacts of peak oil and debt-deleveraging.¹ I propose that scarce means less affordable -i.e. oil price as a percentage of median income*. If wages go down but prices go up slightly to say $90, this has the same scarcity effect as wages going up a bit and prices going to $150. I cannot predict which combination it will be, and I cannot be certain enough of a certain price level at certain time, to make buying oil futures worthwhile. Along with falling wages, I also expect financial instability. Financial instability that could very well lead to the oil market crashing as it did in 2009. Timing is everything.

The Adaptive Market Hypothesis*

We now turn to an alternative explanation of the oil markets and the Adaptive Markets Hypothesis (AMH).  The oil market and information regarding oil supply and demand can be described as VUCA – Volatile, Uncertain, Complex and Ambiguous (yes, that’s a real acronym*). Traders manage a VUCA environment by distilling information into trading strategies, subject to cognitive biases. For example, ’If oil goes below $65 then buy’, or ‘if the market shows this pattern, then sell’.  The AMH brings an evolutionary perspective to how markets function.  Markets…

“…can be seen as a co-evolving ecology of trading strategies.
The strategies would correspond to biological species, and the
total capital commanded by a strategy is analogous to the population of that species. New strategies are constantly created, thereby changing the profitability of pre-existing strategies, in some cases even replacing them or driving them to become extinct.” (Mandelbrot, 2008, p. 72)

Survival is the name of the game, and traders are forced to constantly adapt their strategies or drop out. The AMH offers a better explanation of how markets function. There are three suspect assumptions that underlie the efficient market view which are not a problem for the AMH:

1. Inaccurate beliefs do not dominate the market. Keynes’ famous quip “Markets can remain irrational a lot longer than you and I can remain solvent.” applies here.
2. Traders act as cool calculating agents and are able to weigh up one piece of information against another.
3. Information about oil supply and demand is clear and unambiguous.

In addition, The efficient market view is not contradicted by, but does not sit well with some facts we can observe about the oil market and markets in general:

1. Although the EMH, implies that trying to predict the oil market is futile, the majority of traders use ‘technical analysis’, techniques that explicitly try to predict what the market will do. What gives?
2. Markets often crash. Low frequency events have a large (usually downside) impact on markets. The probability distribution of future crashes is not knowable or computable, and does not conform to the Gaussian distribution* that underlies other economic theories (e.g. Black-Scholes, Portfolio Theory) (Taleb, 2007)  
3. Traders are human, and are subject to all the cognitive biases, and behavioural quirks that psychologists have studied over the years. Including herding -doing it because everyone alse is.

The main takeaway is that while cornucopians present the efficient market view as arising from well-grounded economic theory, it is really just one, non-falsifiable view of how markets operate. Future posts in this series will flesh out these points.

References

Mandelbrot, BB, (2008) Replicating the Stylized Facts of Financial Markets. Lecture Notes in Economic and Mathmatical Systems.

Taleb NN, (2007) The Black Swan: The impact of the highly improbable

1. George Mobus explains this from a big picture perspective here

How Engineers View Peak Oil

This post presents how someone of an engineering bent approaches peak oil. Whereas economists focus, naturally enough, on prices, engineers are more interested in the flows of energy and the technologies that transform energy into services.  From a peaknik engineer’s perspective, peak oil means a decline in liquid fuel flows, and subsequently, the decline in the energy services we get from burning liquid fuels in engines.

First, let’s take a look at flows. The chart below give an historical perspective on crude oil and other liquid fuels over the last century or so. The three other liquid fuels are: Natural Gas Plant Liquids (NGPL), these are heavy gases like propane, butane, ethane, that are distilled from processing natural gas, and are often marketed as LPG (Liquified Petroleum Gas); I’ve aggregated Canadian tar sands, Venezuelan extra heavy, and oil produced from coal and natural gas feedstocks into Unconventional Oil, most of this category is tar sands; finally the green sliver only visible in recent years is Biofuels, most of this is corn and sugarcane ethanol from the USA and Brazil respectively. This chart required a reasonable amount of data wrangling because different data sources aggregate these liquids differently, and report fuels in volumetric (thousand barrels per day) rather than energy equivalance terms (e.g thousand barrels of oil equivalent per day). NGPL and ethanol have about 65% and 60% respectively of the energy content of crude oil per equivalent volume, so volumetric reporting can misleadingly overstate the significance of these fuels. (Data: EIA; BP; Caithamer, 2008*;  Earth Policy Insitute; CAPP*)   

In terms of services, liquid fuels largely power transport, and of that, private cars. Significant amounts still go to industrial uses (as an energy carrier, asphalt, plastics etc.); residential/commercial (a lot of houses in the northeast of the USA use oil for heating); and electricity generation (particularly in countries without reliable electricity grids).

Data: EIA

Oil flows have to through a number of stages before they coverted into services. For simplicity let’s just consider consider transport at this stage. What I’ve been calling services an engineer or a physicist would call work -that is in this case, the motion of the car. To illustrate how flows are transformed into work, let’s follow a barrel of oil in its simplified journey from coming out of the ground to fuelling a car. An oil barrel contains 160 litres of oil or 5.9 Gigajoules of chemical energy. First, the oil needs to be refined into petrol. Refineries consume about 6% of this for their own energy needs, leaving us 5.5 GJ of net petrol. but thanks to the magic of cracking*, and because petrol has less energy per unit of volume, we actually come out with 161 litres of fuel. Combustion in a car’s engine wastes about 63%¹ of the original barrel’s energy as heat, while a further 14% is wasted through the transmission, when the car gets stuck in traffic, and must stop and start (in typical usage). This leaves about 0.95 GJ of mechanical work. That’s still quite a lot. Those of you who remember your physics will recall that 1 Joule is equal to the work required to accelerate a 1kg object 1m/sec . 1 GJ is a billion times this amount. In more familiar terms our 161 litres of petrol will move a toyota corrolla (7.84L/100km) 2050 km assuming typical usage, despite the engine dissapating most of the energy as heat. Completely coincidentally (I swear!) 2050km is the road distance between Cape Reinga and Invercargill*.

Having reviewed the basics of flows and transformations we can now contrast the standard narrative with the peaknik engineer narrative. The following is a year old chart from The Oil Drum that aggregates most current peaknik oil extraction forecasts. The chart is volumetric and shows Crude Oil and Natural Gas Liquids (CO + NGL) combined (often this combination is called conventional oil). A median of all peaknik forcasts has us losing 15% of conventional oil by 2020.

[chart source]

So, taking these analyses at face value and without any mitigation efforts (ceteris paribus as economists like to say) there is a risk we would lose 15% or more of  our oil fuelled energy services. There are four types of efforts that mitigate this:

1. Reduce demand: change the urban form to reduce the need to drive; change out the car fleet to lighter vehicles; telecommute through the internet.
2. Introduce new technologies that do not rely on oil. Electric cars are the main idea here (of course, you still need to power these with electricity but we’ll ignore that for the moment).
3. Improve the efficiency of the internal combustion engine (ICE) (the ICE has been around for over a century so we shouldn’t expect massive gains here).
4. Produce more liquid fuels from plants or other feedstocks.

Peakniks agree that the above four will continue to occur, and at greater rates than present. However, they contend that the rate of crude oil decline is likely to exceed the rate at which these four can substitute for the loss of oil fuelled services. Further, although the peaknik narrative gets a bit fuzzy at this point, that these services are integral to the modern economy.

The standard narrative implies that the peaknik analyses must be wrong; and/or, that the four mitigation efforts will almost seamlessly substitute for oil with little or no loss of services; and/or, that the loss of these services will have little impact on the economy. Peaknik engineers, who are perhaps more familiar with how long it takes to introduce and scale new technologies, are sceptical. The peaknik wishes to know that if oil was to decline in the near-term, then which feasible combinations of biofuels, electric cars etc can substitute for oil, and how fast and how far will can we reasonably expect these to come online? 

The blog project hinges on these two points: 1. There is a risk that conventional oil will decline in the near-term; and 2. that the rate of oil decline could exceed the rate of substitution. If economists can agree we can make progress.  If economist disagree and believe these to be impossible, then I expect justification.

Further posts will develop these points and I will switch back and forth between economic theory, and a feasibility assessment. The feasibility assessment will attempt to answer the following -When can we expect conventional oil decline and how fast will this occur (with an uncertainty range)? How fast can mitigation efforts reasonably substitute for oil’s decline? Mitigation efforts 2-4 are quantifiable and commeasurable with a decline in oil extraction. Urban form and the like are a bit trickier and I will leave those to the end. To round off, I will attempt to determine what affect a loss of oil-fuelled services could have on the economy. The next post will look at the Efficient Market Hypothesis, as applied to oil markets and following that I’ll formalise the feasiibility assessment and look at oil discoveries.

 1. Efficiency data taken from  Robert U. Ayres, Leslie W. Ayres and Benjamin Warr (2003). Exergy, power and work in the US economy, 1900–1998, Energy Vol 28, Issue 3 http://dx.doi.org/10.1016/S0360-5442(02)00089-0

Posts in this series:

Blog Project: The Economics of Peak Oil

How Economists View Peak Oil

Petrol Affordability

I’ll explain how I made these later and upload the excel file.

How Economists View Peak Oil

This post outlines how economists typically think about peak oil,  what I call the ‘standard narrative’ of resource transition. The purpose of this post is to lay the groundwork for a more constructive debate regarding our understanding of peak oil. Economists are generally less concerned about peak oil than the geologists, physicists and engineers of ASPO (see sidebar). In this post I will attempt to identify the components of economic theory that support their view and how they weave these components together into a routine resource transition narrative.

Of course, not all economists share the standard narrative perspective. Here, I use ‘economist’ as a shorthand for ‘orthodox economist’. The much less numerous biophysical economists (see sidebar), are economists and are pretty concerned.  But since biophysical economists reject a lot of orthodox theory, and have a backgrounds in physics and ecology, they are economists in about the same sense that the Jehovah’s Witnesses are Christian.

There are a few orthodox economists that take peak oil seriously. Jeff Rubin, recently Chief Economist at CIBC World Markets, is clearly in the ‘peaknik’ camp. Rubin is a speaker at the recent ASPO-USA conference in Washington, DC.  Rubin argues that peak oil will involve a painful transition to a less globalised world. Somewhere between the ‘peaknik’ (more concerned) and ‘cornucopian’ (less concerned) camps lies James Hamilton. Hamilton is an expert on oil shocks and has identified the steep preceding rise in oil prices as a key factor in US recessions (including the most recent one). Do a search through the econbrowser archives (sidebar) and its pretty clear that Hamilton considers peak oil to be a major problem, but perhaps wouldn’t go as far as Rubin.

I will draw together a standard narrative by combining and condensing the following analyses with a focus on material published by NZ economists:

How to talk to an economist about peak oil -by James Hamilton (publ. 2005).

What happens after oil production peaks (2005) and Energy Futures: An economic framework and some policy implications (2006) -both by Brian Easton.

Oil price assumptions for energy outlook (2005) -by Ralph Samuelson (Ministry of Economic Development) and Michael Taylor; and Oil: An introduction for all New Zealanders (2008) -by Ralph Samuelson.

What price peak oil? (2008) -by Gareth Kiernan (Infometrics).

The latter two articles by Easton, the article by Samuelson and Taylor and the article by Kiernan are the only analyses of peak oil, authored by NZ economists, that I have been able to find (apart from the odd blog post). A situation which  should raise concerns in itself. If you know of others please let me know.

Economists focus their analysis on oil prices, classifying peak oil as a problem of increasing oil prices. The standard narrative counters a view that prices will rise rapidly and stay permanently high, instead arguing that prices will rise moderately and be moderated by improving technology. There are five parts to the standard narrative explained below.

 

1. The oil market signals increasing scarcity prior to the peak through moderate price increases.

Here economists are invoking the Efficient Market Hypothesis (EMH). Efficiency in the EMH sense means efficient at processing information. The idea is that because there is:

(a) an oil futures market where traders trade contracts for the delivery of oil out to eight years (see Mazama Science on the sidebar for data from the oil futures market); and

(b) oil traders have the incentive to make a profit and anticipate moves in oil prices by buying low and selling high; then

accordingly, the numerous oil traders act as vast information processing mechanism. Because traders have the incentive to discover and incorporate information about oil supply and demand into a trading strategy, the price of oil should reflect all public information about oil. We can  reasonably expect the oil market to anticipate any scarcity issues. From a public policy point of view, the EMH suggests that because over the long-run a single forecaster can do no better than the futures market at anticipating prices, governments should allow energy markets to allocate investment, and should not ‘pick winners’ by intervening with large-scale public investment.

 

2. Anticipating further prices increases, oil extractors defer extraction.

If an oil company is confident that oil prices are going to go up, then the company can make more money by waiting for the price to go up and extracting the oil later. Economists use the observation that oil companies are not deferring extraction* as evidence that peak oil (here classified as a rapid price increase) is not a near-term problem and/or will not result in permanently elevated prices.

The above conclusion stems from the Hotelling Rule*, named after the economist Harold Hotelling who’s 1931 paper, The Economics of Exhaustible Resources is probably the most cited resource economics paper ever. Resource economists have attempted to determine the most profitable extraction strategy, and price evolution under different conditions.   The maths are more complicated than I will go into here, but basically Hotelling finds that the assuming complete knowledge, the price of a resource should rise in line with interest rates (interest rates influence this result because oil companies can invest the profits of extraction into an alternative interest bearing alternative). The observation that historically, oil prices have generally fallen (see chart below) is used as evidence that technological improvement has overcome scarcity in the oil price ‘tug of war’ (Samuelson, 2008). According to the standard narrative, we can expect technological improvement to continue to moderate prices.

 

3.  As prices rise, oil reserves and resources that could not be extracted profitably, become profitable.

Deepwater oil, Canadian tarsands, Venezuelan extra heavy and oil shale are not profitable to extract at $US20/barrel but become profitable at perhaps $US75/barrel or higher. There are vast amounts of unconventional oil resources. According to the standard narrative, we can expect a rise in prices to boost extraction rates above a  peaknik analysis. A rise in prices also encourages discovery efforts and we can expect to discover more oil.

 

4. As prices rise alternative technologies become profitable, accelerating a transition to alternative sources of energy.

As for 3 above but for alternative technologies such as electric cars, biofuels, and coal-to-liquid plants. Economist William Nordhaus outlined the theory behind this in 1973. Nordhaus called alternatives that can substitute for nonrenewable resources ‘backstop technology’. Nordhaus argues that because the potential of underexploited and more expensive energy resources such as solar and nuclear are so large, we do not have to worry about spiralling energy prices. high prices create a market opportunity that  investors fill by starting alternative energy industries that substitute for oil (electric cars) or provide alternative fuel (biofuels, coal-to-liquid). Historically, the price of alternatives has fallen with improving technology. The (falling) price of substitutes will therefore cap the price we pay for energy services (see chart below from Samuelson and Taylor, 2005). Electric cars, such as the Nissan Leaf, are entering the market  and demonstrate this principle in action.

 

This chart is from Samuelson and Taylor’s article on the oil price assumptions of the 2005 NZ Energy Outlook. The High Oil Price case represents what Samuelson and Taylor believe could happen during near-term peak oil. A transitional period of elevated oil prices 2010-2015, followed by a fall back to $US75/barrel as alternatives come on stream.

5. Oil consumers cut consumption of oil by substituting away from oil. Through substitution and efficiency the economy becomes less dependent on oil.  

The standard narrative is often rounded off by observing that while both oil consumption and GDP have grown, oil consumption has not grown as fast as GDP. Economists argue that the economy is becoming less dependent on oil. According to the standard narrative, we can expect this process to accelerate with higher oil prices. For example, the chart below presents NZ real GDP, oil consumption, oil prices (right axis); and GDP per barrel of oil consumption (left axis), a measure of oil efficiency. High oil prices coincide with an increasing ratio between GDP and oil consumption. GDP may grow at a slower rate through the transitional period of high prices.

 

Data: Statistics NZ, BP, Penn World Tables.

 

6. Summary

Prices rise preemptively and moderately, signalling the market opportunity of alternative investment. People respond to incentives: finding more oil, innovating and bringing alternatives to market, and reducing consumption of oil through substitution. We may experience a transitional period of slower growth but the economy adjusts.

 

 Conclusion

Reassured? While I think that the presented theories contain important insights, I do not find the standard narrative reassuring. Following posts will outline its deficiencies, and offer alternative economic and ‘engineering/physicist’ frameworks for understanding peak oil. At this stage, we should note that the standard narrative does not offer much in the way of recent real world data, other than prices, or any demonstration of feasibility.

I am emailing a number of economists and other interested parties to make them aware of this project and to invite criticism, commentary, and error correction.  Internet-based dicussion offers ‘instant peer review’ and  could attract a wider audience than the traditional journal process. Perhaps this project will eventually be written up for publication in a journal. Perhaps my promotion of limits to growth considerations will be dismissed out of hand as the ravings of someone with no qualifications in economics! In any case I will push on. The following post will look at how an engineer might approach peak oil and after that I will shift back and forth between economic theory, and an ‘engineering feasibility’ analysis.  

 

P.S. The economic theory posts will run along the following lines:

1. Adaptive markets; uncertainty; allocative and informational efficiencies.

2. Prices and costs as indicators of scarcity.

3. Liebig’s law of the minimum and the logistics of alternatives development.

4. Energy Return on Invested (EROI) and the ‘receding horizons’ problem.

5. Critical dependence on oil? Reconciling biophysical and orthodox economics.

 

 

 

New Zealand Debt Deleveraging

Since I already have charts made on NZ debt levels and debt deleveraging seems to be affecting the economy more than peak oil I thought I’d take a quick diversion. The weak recovery has surprised economists. A few months ago most were anticipating consumers to spend more before the GST increase. Here’s Shamubeel Eaqub from the NZIER:

“There was no evidence of a pre-GST [increase] spend up at shops. Large firms, which had been recovering strongly, fell sharply, and small firms remained in the doldrums.

“Business profitability is deteriorating again; highly unusual for this stage of the recovery. This may weigh on future hiring and investment although hiring and investment intentions remain encouragingly resilient.”

While many commodity prices – in particular dairy – continue to boom, the domestic economy remains weak as farmers and households focus on paying down debt, even before the recent changes to the tax system, which were designed to encourage saving.

Mr Eaqub said the move to pay down household and corporate debt was hurting the economy at present, but the process would leave the economy in a better position.”*

Paying down debt? According to the consensus forecasts¹ from a few months back, consumers were supposed to spending again. This raises the question of how high debt levels affect the economy (and whether orthodox economics has a good handle on this). Bernard Hickey has been beating the debt-deleveraging drum for a while³. While Matt Nolan has called for a thorough look at the data².

I’ll present NZ debt data further down but before getting to that we should note that across the Tasman, University of NSW Western Sydney Professor Steve Keen has done a lot of work on debt levels. Keen is in the Post-Keynesian school of economics and rejects a lot of orthodox theory, having written the book Debunking Economics. Keen has a pretty clear explanation for the cause of the financial crisis: Too much debt.  Keen regards the current situation as comparable to the Great Depression. Here’s a chart of private debt levels in the USA, then and now I poached from Keen’s post*.

Looks somewhat ominous doesn’t it? In Keen’s view, aggregate demand is GDP + the change in debt, an approach which recognises that asset markets and consumption are mixed. Here’s the USA’s aggregate demand:

And here’s the correlation with unemployment:

Keen points out that change in debt is the leads change in employment while gdp lags. I recommend reading the rest of Keen’s post and The Roving Cavalier’s of Credit* for the theoretical background. The rest of this post merely presents NZ data from the RBNZ in the same framework.

Total debt ratios are comparable to levels preceding the Great Depression. Businesses and consumer debt have been deleveraging for a while now, while agriculture and mortgage debt have slowed down. A rise in government debt has only barely offset a decrease in private debt. Let’s look at aggregate demand:

 

Right around the start of 2008 aggregate demand has fallen, slowed down by an increase in government debt.

Now let’s take a look at the correlation with unemployment:

Looks to me like debt deleveraging explains current economic weakness well. If debt levels reset around 50-100% of GDP, that’s going to take a decade or two. The obvious question is why orthodox economics have missed the role of debt accumulation until quite recently

1 NZIER Consensus Forecasts

2 Brief Thoughts on NZ’s Debt Position

3 Opinion: RBNZ’s more modest rates outlook shows how debt deleveraging has changed the world

Blog Project: The Economics of Peak Oil

I’m neither an economist or an engineer but I’ve spoken to people from both disciplines about peak oil and I can’t help but notice that these two have differing views on the impacts of peak oil. Generally speaking, economists see peak oil as the end of cheap oil and a transition to alternative liquid fuels and transport technologies. Engineers see peak oil, to put it simply, as less liquid fuels = less transportation (= less economic activity?).

Economists tell what I call the ‘end of cheap oil narrative.’ It goes something like this: ”the increasing scarcity of oil results in higher prices. On the demand side, higher prices encourage motorists to reduce consumption and substitute away from oil, adopting more efficient cars etc. On the supply side, investors realise that money can be made from enterprises that produce alternative fuels and substitutes. An alternative fuel technology, such as coal liquefaction, might cost $US70/barrel to produce. This technology was previously unprofitable but becomes profitable once the oil prices exceed $US70. Investment flows into these technologies as entrepreneurs seek out profit opportunities. As these enterprises start-up, the supply of liquid fuels rises, and thanks to competition, prices stabilise just above the costs of the marginal producer.” In the economist view, the higher oil prices around the peak might result in slower economic growth, but in general the world adapts and the economy ticks along as usual.

The engineers’ and peakniks’ narrative might go something like the following or though you could find others: “With an accelerating search effort for dispersed volumes the discovery rate for oil fields follows a bell shaped curve which has peaked in 1960. The rate of oil extraction lags this by several decades and is set go into a terminal decline in the near future. Substitutes for oil do not come online at the scale and speed necessary to replace oil. Oil becomes steadily less affordable, with a volatile prices reacting to geopolitical events and economic disruptions. The efficiency of vehicles improves but this occurs slower than total liquid fuel declines. Liquid fuel powered activity (i.e. most transportation) declines as the rate of oil extraction declines. It is difficult to gauge precisely how a decline in transportation will impact the economy but since most goods have long production and transportation chains, then this could plausibly have a severe impact.”

Economists tend to focus on prices, while engineers focus on physical flows of energy. Now, I’m with the engineers on this one. I think peak oil is going to more serious than the standard economic narrative. That said, economics can still contribute some insights. The problem is that the vast majority of public policy professionals have some training in economics and are therefore much more familiar with the economic narrative. While few have any training in engineering or physics, or similar disciplines and are perhaps less comfortable with the engineering narrative. When the government eventually responds/reacts to peak oil there is a risk that poor decisions will result. This blog is an attempt to build up the knowledge base in an accessible way. It will attempt to engage actual economists and engineers and try and work through the problem from both ends. The following outlines how I expect this will run. Traditionally, you’re supposed to state your opponent’s arguments in the strongest possible terms and then work through the weaknesses, so we’ll start with a mainstream economic perspective on peak oil.

1. The Economists’ Narrative: The End of Cheap Oil

2. The Engineers’ Narrative: Reduced Flows of Liquid Fuels

• How much oil?
• Oil Discoveries
• Conventional oil decline rate
• Tar sands
• Venezuelan heavy
• Coal to liquids
• Gas to liquids
• Shale and other liquids.
• Biofuels
• Transport substitution, efficiency improvements and conclusion.

3. Compare and Contrast

Economic Theory 1: The Efficient Market Hypothesis

Economic Theory 2: The Adaptive Market Hypothesis

Economic Theory 3 : Backstop Technology Substitution

(Biophysical) Economic Theory 4: Energy Return on Invested and Receding Horizons

Economic Theory 5: The Hotelling-Hartwick Rule

Economic Theory 6: Noorgard-Reynold’s critique.

Economic Theory Conclusion.