Refining the Plateau

Year-on-year change in refinery capacity 1965-2005 by two different estimates. Click to enlarge. Sources: BP Statistical Review of World Energy 2005, and EIA Crude Oil Distillation Capacity (Table 36).

Let's continue with the discussion of the production plateau since last year that we first discussed at Thanksgiving, and then resumed just before Christmas. If you aren't sick of it yet, this is the phenomena we are seeking to explain:

Average monthly oil production from various estimates. Click to enlarge. Believed to be all liquids. Graph is not zero-scaled. Source: IEA, and EIA. The IEA raw line is what they initially state each month. The IEA corrected line is calculated from the month-on-month production change quoted the following month.

Note the leveling off that began last summer. Our task is to understand why that occurred in the face of a rapidly growing world economy and ever higher prices for several years, which our economist friends would tell us should be calling forth more and more supply. Are we at, or almost at, the peak, is there some more innocent explanation?

The ever-optimistic Freddy Hutter suggested that

Peaksters can get excited all they want watching the mid 2005 plateau and read into it what they want but the rest of us know it was hurricane and refinery related and the trendline will continue into 2010 amid the gnashing of teeth.
Now yesterday, I think we put the hurricane explanation out of it's misery. So that leaves the refinery capacity idea. Of course Freddy is not alone in proposing that story. As Alexander's Oil and Gas Connections reported
Saudi Arabia's Minister of Petroleum and Mineral Resources Ali al-Naimi blamed lack of refining capacity to handle sour and heavy crude for the spiralling global prices. "What the global oil industry confronts today is a challenge of deliverability," said al-Naimi addressing a packed session [...] "This is because there is a major constraint in the refining system. There is a mismatch between the configuration of refineries and availability of sour and heavy crude," he pointed out.
Firstly, let's orient ourselves to the overall history of refinery capacity. Alas, I was not able to find a monthly series for global refinery capacity, but I found two annual series, and I think they definitely shed some interesting light on the question. One is from the BP Statistical Review of World Energy 2005, and the other is from the EIA. As usual in this game, the data from different sources agree in the big picture, but not on the details. Here's the story on capacity:

Refinery capacity 1965-2005 by two different estimates. Click to enlarge. Sources: BP Statistical Review of World Energy 2005, and EIA Crude Oil Distillation Capacity (Table 36).

Before we compare capacity to production, I would just like you to note how rapidly capacity grow in the late sixties and seventies. I would also draw your attention to the difference in slope between the late nineties, and the anemic climb of the last few years. We will return to these points in a few graphs. First though, how much of this capacity was being utilized?

It turns out not to be easy to compare EIA production numbers with their refinery capacity, so I stuck to the BP numbers for this next bit. Here's a graph showing how much of refinery capacity each year was actually taken up with production.

Refinery production and spare capacity 1965-2004. Click to enlarge. Source: BP Statistical Review of World Energy 2005.

We can see that things are starting to get a bit tight at the end there. However, if we look at the ratio of production as a proportion of refinery capacity, we see that it's no tighter in recent years than it was in the 1960s, when neither oil nor gasoline were particularly expensive.

Refinery utilization 1965-2004. Click to enlarge. Source: BP Statistical Review of World Energy 2005.

So we might wonder why the big fuss now? To me, the most revealing graph is this one. This shows the year-on-year growth rate in refinery capacity.

Year-on-year change in refinery capacity 1965-2005 by two different estimates. Click to enlarge. Sources: BP Statistical Review of World Energy 2005, and EIA Crude Oil Distillation Capacity (Table 36).

Points to note:

  1. In the late sixties and seventies, we knew how to grow global refinery capacity by 6%-10% annually.
  2. As recently as the late nineties, we were growing global refinery capacity at 2% a year.
  3. On several occasions, we have increased the growth rate in refinery capacity by one or two percentage points within the space of a year or two. Eg in the late eighties, and again around 1993-1994.
  4. The only period of near-constant very slow, nay positively anemic, only 1/2%, growth in refinery capacity is the period 2001-2004.
  5. If we had grown refinery capacity at 1%-2% from 2002 on, there would have been no problem with refinery capacity. The problem was not just that demand grew, it was also that refinery capacity growth was very minimal during that period.
Just how fast was production growth in the run-up to the plateau? Here it is:

Year-on-year change in monthly EIA oil production Jan 2002-October 2005. The dark green line is the data, and the plum curve is a quadratic fit to the data to illustrate the smooth trend. Click to enlarge. Source: EIA.

So demand growth certainly became respectable -- getting up to about 4% in the smoothed trend -- but nothing out of the historical norm, and generally inline with GDP growth in a healthy economic recovery. Given all this, I cannot believe that we just couldn't increase refining capacity in time to avoid the bottleneck this year. On the historical record, we certainly should have been able to get it up to 2% annually, or even a little better, by now, and that would have been enough to avoid the bottleneck. No, it's not that we couldn't, it's that we didn't. We chose not to.

I can think of only two reasons why we might not have chosen to increase refinery capacity in the way that we obviously could. One is a secret refining cartel. The other is peak oil. I think Calculated Risk made the key argument back in the summer. If the problem was just a refinery cartel, crude oil, or at the very least heavy grades of crude oil, would be selling for $10/barrel. It's not - even the heavy stuff is $50/barrel. That implies there's very little of it to spare, not that there's quite a bit but it's not making it through the refining cartel.

What it looks to me like is the refinery market, explicitly or implicitly, understands that ramping up to compete for market share on CERA's 100mbpd+ of production in the near future would be a very dumb business move. Instead, it looks like they are inching their capacity up in the most gingerly manner possible. They're acting like their big fear is getting caught with too much capacity.

So I call bullshit on this refinery explanation.

Anybody got any other ideas for why this plateau might have happened besides imminent peak oil?

There probably is a lack of refining capacity for heavier crudes, and supplies consists of an ever growing part of these.

I really do think we are presently at Peak Oil, and this may be reflected in the refinery market where PO has been anticipated by not expanding capacity.

If the above is true, your analysis raises another important question. How long will we be able to maintain the plateau, especially in the light of heavier crude entering the market?

Here is an indication that more and more heavier crude is in the market
If there was an ample supply of sour crude that was not being refined it would be offered for much less than $50.  It is not credible that the price would go down if sufficient sour crude refining capacity came on stream.  In Saudi Arabia sour crude comes from small fields that could be called leftovers.  There isn't a sour crude Ghawar.  In Russia, sour crude comes from old depleting fields, mainly in Tatarstan.  There is not a vast amount of sour crude available in contrast to totally degraded oil found in tar sands.
That's right.

The spread between light and heavy oils has been growing though. While I think this is primarily because of decreases in available light sweet crudes, refinery configuration and capacity do play a complementary role.

According to Matt Simmons, the Safaniyah offshore field in Saudi Arabia produces heavy crude, and that is their largest field after Ghawar.
Thanks again for your work!  Do we know if, worldwide, the overall expansion is new refineries, or expansion of existing refineries? I hear building a new refinery is 4 or 5 years, whereas if capacity can be increased quicker at an existing one, and can account for the 1-2%( at existing sites) then this might give some idea of how the industry knows PEAK, or is just making an individual business guess and the collective decisions add up to what you  quantify for us today.    Also in the states I read we haven't built a refinery since the 70's. I guess the declining production  is the real reason, not the NIMBY/cost factor often given?
Looking at refinery sector development only in terms of the volume of crude facilities take in misses the real story of refinery development in recent years. "Capacity" basically refers to the amount of oil a facility can run through its distillation columns, the first stage in refining. It does not speak to the quantity of product produced or the amount of investment in the sector.

These factors are better explained by measures of complexity and conversion, which refer to the ability of a given facility to produce valuable products. Complexity describes the stages of refining that occur after distillation, primarily cracking. These stages are capital intense, actually transform the molecular structure of crude and significantly alter the output of a facility.

If you were to measure the amount of gasoline produced and/or the dollar value of investment in the sector, you would get a completely different result. It is inaccurate to claim that refinery development ground to a stop in the 1990s.

The main reasons why new capacity did not come on line in the recent past are:

  1. Adequate crude intake capacity
  2. Ability to produce more product from other investments in the facility
  3. Regulations that made new facility development difficult, and most importantly,
  4. Refining was not a profitable industry and did not proivide acceptable returns on investment.

I have been in discussions with bankers and refining executives and seen huge amounts of analysis. If fear peak oil is a signficant factor limiting development, it is the world's best kept secret. Additionally, there several huge new investments in additional refining capacity taking place in India and China, partially funded through int'l investment. Surely the same logic has to apply there.

It do think fear of peak oil could be an issue in the future, but see little evidence it has in the past.

If anyone wants to look at a presentation to investors for a very typical facility upgrade, download the 2+ MB PP presentation below.

Bangchak Petroleum is a simple refinery in Thailand that currently consist of distillation and some basic postdistillation treatment (incl. noncatalytic cracking). This configuration is a standard simple refinery like that found in the US in the past.

The plan to invest $250 million to become a complex refinery. There will be no increase in "capacity", but the new facility will produce much more valuable product (gasoline and diesel) and much less low value product (fuel oil).

This is a very good example of how the modern refining sector is investing and developing. The presentation is fairly clear and provides a great insight into why refining is not as clear as it appears.

If you don't want a lot of (mostly useful) background, skip to slide 25 which describes the upgrade calling it a Product Quality Improvement Project.

I'm not understanding something - are you are saying these investments will result in more complex products from the same input of crude, or more output (i.e. are they increasing the efficiency of refining)?  Also, how is diesel so much more complex that heating oil (I run heating oil in my tractor!)?

From my admittedly ignorant perspective, I find it hard to swallow that refineries were not already pretty good at making diesel and gasoline.  What kinds of throughput improvements are possible for these outputs?  And if the investments are to make other more profitable but lower volume distillates, then does it really matter?


Diesel and heating oil are basically the same until the very end of the refining proces where they are treated differently. Fuel oil is a far lower quality product and is the only significant output that is worth less than crude (see Bangchak slide #12).

A complex refinery does produce a lot more gasoline and diesel from the same barrel of oil. Slide 25 of the Bangchak presentation shows that the refinery currently produces 37% fuel oil, 37% diesel and 15% gasoline. The addition of a $250 -350 million dollar hydrocracker and other equipment changes this significantly. Afterwards it would produce 9% fuel oil, 52% diesel and 25% gasoline from roughly the same crude input.

Refineries have always been good at making gasoline and diesel, they just made less of it from the same input

According to the American Petroleum Association, the typical US refiner produced the following products from a typical barrel of oil.

Distillate: 9.7 gallons
Kerosene-type jet fuel: 4.3
Fuel oil: 2.0
Liquefied gases: 1.9
Still gas: 1.9
Coke: 1.9
Asphalt: 1.4
Petrochemical feedstock: 1.1
Lubricants: 0.5
Kerosene: 0.2
Other: 0.4


"...the refinery currently produces 37% fuel oil, 37% diesel and 15% gasoline." should read 31% fuel oil, not 37%.

OK, so then presumably the investment can result in producing more of the outputs people want from the same crude input, so it's probably an attractive one.  BUT, that's different than investing in refineries or modifications that result in more usage of crude.  So it may be that your point actually reinforces what Stuart is saying, in that the investment that is occurring in refineries is not directed towards projects that would increase the use of crude, but toward maximizing the profitability of the existing crude inputs.  And this would make sense if one did not expect larger quantities of crude to be available.
Two points:
1 It is true that no US refineries have been built in 25 years, but this is very misleading - US refinery output, measured in barrels input, has increased around 30% in the period, albeit almost nil since 2000.
2 Refinery profit is dependent on cost of crude less price of product. If crude is rising faster than product prices, as has been true for most of the period from 1998, refinery profits are declining, reducing interest in new investments. (European supplies of product post K/R has naturally cut product prices and refinery profits.)  The only company significantly expanding capacity in the US is Valero, which buys up old, antiquated refineries, adds substantial new equipment designed to handle cheap sour/heavy crudes, and ends up with higher output. So, most integrated oils are not interested in expanding because their profits are anemic, but Valero is happily exploiting its niche. Eventually, if refinery capacity is a true bottleneck, crude stocks will continue to rise (as they have been for the last year) while product prices rise, increasing refinery profits and renewing interest in expansion. OTOH, OPEC may defend $55 NYMEX, restricting crude to match refinery capacity.
Another very good research by Stuart. It is compelling; you would expect the refining business to be aware if there is going to be any serious grow in oil-production. The refinery with spare capacity would be the one that can make extra money. But then it is extra interesting to see that China is not planning on building one but two refineries in the near future. One wonders where this extra oil is going to come from. Or they have other plans.....
Well, the main plateau started in about 1980, and the decline to it started with the oil shock of 1973. It looks like up to that point refining companies had a particular model of demand growth and they just kept building at a certain rate, barely keeping pace. From 1973 on for several years, from your graphs, they must have been really hurting as new capacity came on line that there was no need of.

I grew up in Newfoundland, Canada, and we were very aware of the status of the "Come By Chance" oil refinery, completed in 1974/1975, and then mothballed after about a year of unprofitable operation. It was started up again in 1987 and is fully operational now, but anybody aware of the history would be very reluctant to build a new refinery without being certain we won't get a repeat.

So the particular reason for the decline in refinery capacity additions in the last few years could be expectation of near-term oil price spikes; given that the recent flatness (2001-2004) is coincident with the Bush administration's tenure, maybe oil industry executives knew to expect something... But we wouldn't want to get into conspiracy theories...

In the 1970s, the US removed incentives for smaller refineries. I will find the name and date. Once these were gone some inefficient facilities were shut down, so the nation's capacity shrunk.

For many years after, it was easier an more effective to add capacity further along the process than distillation, this is called debottlenecking. For this reason there have been no greenfield projects in the US. I understand that most opportunities have been exhausted.

But there has been a lot of capital invested in refining and the current infrastructure is not the same as 30 years ago. Demand growth has been for highly refined products such as gasoline, diesel, jet fuel and heating oil. This has been met by additions of crackers and other equipment that create more of these products from the same distillation capacity.

I don't think there is anything conspiratorial about the oil industry waiting on the Bush Administration to subsidize or build refining capacity. I remember the president offering to build oil refineries on military bases several years ago, which I took at the time as a tacit acknowledgement of Peak Oil.
I just received notice of a conference dedicated to a somehwaht parallel question on the natural gas side, that of natural gas storage ( The teaser paragraph is as follows (sorry I can't put it in the cool box that others use for quotes):

A new era in gas storage is dawning in 2006. The passage of the Energy Act of 2005 has provided FERC with broad new powers to drive storage expansion. They have responded by proposing new rules for market-rates for interstate gas storage to spur the development of storage capacity. Hurricanes Katrina and Wilma also demonstrated the value of gas storage facilities, and have caused many players in the gas market to step up their efforts to develop or acquire more storage capacity. High gas prices, increasing volatility, and increasing liquidity of gas markets have increased the intrinsic value of storage facilities to both active market participants and investors, driving an unprecedented round of acquisition and aggregation plays. Likewise, the introduction of new LNG and pipeline sources is likely to change the potential value of existing assets and drive the development of new storage facilities. Gas storage operators, pipeline operators, gas customers, and storage investors must examine the changes in the regulatory and market landscapes to form effective strategies to profit from opportunities in 2006 and beyond.

It seems that the natural gas industry is not running scared to make investments because of potential supply constraints.

I'm not taking any sides!

The general thrust of the argument about (non peak oil) refining capacity is this:

Margins were very bad in the 80s & 90s, therefore no capacity was added. In fact in the USA around 0.9mbpd was closed down.

The oil industry/markets were complacent about demand growth.

Complex refineries are far less than simple refineries in number; hence discounted OPEC heavy/sour crudes.

Now a lot of refineries are being built in the middle east, nigeria, indonesia and china to refine product.

In addition Valero, Conoco & others (i dont have my notes!) are building additional capacity in the USA and europe in existing refinery complexes.

The catch up will come with the next recession, by 2010 when some years of slow demand will allow complex refineries to be built to handle heavier grades.

That's the argument, i'm not saying its mine so dont flame me!!

Here's a recent piece I did: Refinery capacity to test crude price in 2006

all the best

I do not know who invented idea that lack of refinery capacity increase crude oil prices. I consider is completely wrong.
If there is problem with refinery capacity,then oil will be cheap ( who needs it if it cannot be refined ) and gazoline expensive. We have seen this situation after harricanes. Extra oil from SPR was not sold, becuase refineries were unable to process it. As result gazoline was at $3 per gallon, while oil was at $65, normally for gozoline to be at $3, oil should be at least $85 per barrel.
There is lack of refineries for heavy crude and we see the it is sold at discount ( $10 to $13 per barrel discount for heavy crude ). Once new refineries of heavy crude is build, then the price of heavy crude will go up, there will be more demand for it.
Yes it seems a bit wrong headed that a lack of capacity in refining drives up crude. But it does tend to in most cases.

Basically if gasoline is $3 there is a need to get more crude through the complex to lower the price.

That drives up crude prices, that is caused by the `bottleneck` and is analagous to the discounted heavy crude no one wants...

If there were lots of complex refineries it could go either way, too far and prices would fall (spare capacity) and too little and the same problem remains... seems a bit wrong headed that a lack of capacity in refining drives up crude. But it does tend to in most cases.

I believe your argument is unsound, and the problem with it can be illustrated by the extreme case.  Suppose there is hypothetical disaster that removes all petroleum product refineries except one on the U.S. Gulf Coast, whose entire capacity can be satisfied from current Gulf of Mexico production.  Would not the world gasoline price be astronomical and would not Saudi Arabian crude oil for export would be worthless?
I think there is a disconnect, but it is temporary. As product prices increase, crude "foolishly" follows it up until stocks climb to the point that nobody wants more crude, then prices decline, which is exactly what is happening right now. However, OPEC is unhappily watching the decline, which is why they are likely to reduce output at the end of the month - so, once again, product prices can pull crude up with them. Actually, OPEC is probably itching to show they have control over the market, lost for the past three years.
Hi Slippery
Yes like i said it does sound a bit odd. But if the case arose as you put it the world would be screaming for gasoline and you need crude to make it. But i think your case is a bit extreme to make a really definitive prediction.

The prices of nat gas, crude, gasoline, heating oil and even emissions credits in the EU tend to drag each other around quite a bit.

The markets see that simple (light sweet) refineries can't cope (if the price of gasoline is high) so this forces up the price of light crudes. At the same time, analagously, heavy crude is sold cheap because there are not enough refineries to put it through the system to make the products the market wants.

I'm not putting it forward as my argument btw, thats just the way the market reacts.

If your case arose it would be very very extreme, you could see all sorts of trades going on like huge longs and shorts on products and on their crack spreads.

At the moment that's kind of the way the market gets it, why refinery bottlenecks affect price, why they are building so many new refineries now (in the m/east)..

The problem I have with this argument is that heavy crude isn't being sold cheap. While it's true that spreads have increased, heavy crude is now much more expensive than light sweet was three or four years ago. Again, this graph Dave posted shows the recent history:
Brilliant analysis, Stuart. This goes back to a statement I've made previously, that you can see what the IOCs really believe by following the money. The money is not in more exploration. The money is not in more refinery capacity. The money is going into gobbling up smaller players in order to keep their own reserves stable (but not even growing). Your analysis confirms what the money tells us - the IOCs do not believe that investing more in exploration, processing, or production are profitable ventures.
As indicated by the two posters above (Adam and Jack), it seems that there is not enough refinery capacity to handle heavier/sour crude especially in the US. A few weeks ago, a EIA weekly petroleum status report was reporting that refiners seemed not to use their capacity at the fullest despite very good margins which could be an indication that the fraction of heavy crude within the inventories was higher than usual. Basically, simple refineries designed to handle light sweet crude need to be upgraded toward more complex ones. But this transition won't necesseraly show up in the overall refinery capacity numbers. It could be very interesting to look at the evolution in time of the fraction of the world refinery capacity that can handle heavy sour crudes but I don't know if that kind of statistics is available anywhere. There are also big differences among countries, for instance South Korea has a refinery infrastructure that has been designed to handle mainly heavy sour crudes which constitute the majority of their imports.
"Basically, simple refineries designed to handle light sweet crude need to be upgraded toward more complex ones. But this transition won't necesseraly show up in the overall refinery capacity numbers. It could be very interesting to look at the evolution in time of the fraction of the world refinery capacity that can handle heavy sour crudes but I don't know if that kind of statistics is available anywhere."

Right. A lot of this has already happened with the driver in the past being demand for products rather than need to process lower quality crudes. It may be hard to distinguish between the two investments. The most useful statistics may be refinery capacity, capacity utilization, complexity and conversion by country. Complexity and conversion are measures of the ability of a refinery to produce better products from worse inputs - creating higher refining margins.

I think this is the right question and you would find that in the US much of the potential upgrading has been done. The US does not have any major simple refineries and has been doing more sophisticated debottlenecking. I think there is more low hanging fruit in other countries.

In a year that has seen oil prices reach record highs, it may seem odd that producers have been offering discounts to get rid of the stuff. But that has been happening with crude oil known as "heavy sour," which is different from the "light sweet crude" whose per-barrel price is most often quoted as the price of oil.
In fact, more than three-fourths of US refinery capacity can process heavy sour, which typically sells for a few dollars less than light sweet crude because it is not as easily refined. And this year, as Persian Gulf producers have flooded the market with additional supplies of the heavier sour crude, the sweet and sour price gap has grown even wider, reaching $ 17 to $ 18 the last week of October.

That's good news for US refiners, which are considered to be in a better position than those in other countries to take advantage of the discounts. Indeed, earnings announcements by US refiners are singling out the cheaper sour crude as a major reason for their growing profits.
"This has resulted in excellent refining margins," said Curtis Hyatt, an associate director for the Cambridge Energy Research Associates, an energy consulting firm. But it's unclear how much, if any, of the savings from the cheaper crude oil have benefited US consumers.
I suspect the corporations short time horizon has a lot to do with it.  The complex refineries cost billions and take 5 years to build and another 5 years to get pay back.  Why invest long term if it might hurt your bonuses?
Quote from a 2001 report on refinery capacity by Senator Ron Wyden (Oregon):

However, the record shows - supported by documents I have obtained - that there is more to the story.  Specifically, the documents suggest that major oil companies pursued efforts to curtail refinery capacity as a strategy for improving profit margins; that competing oil companies worked together to subvert supply; that refinery closures inhibited supply; and that oil companies are reaping record profits, yet may benefit from a proposed national energy policy that would offer financial incentives to expand refinery capacity.

That last bit rings true. Here's the industry spokesman in September 05:

Bob Slaughter, the president of the National Petrochemical and Refiners Association, told a House committee last week that Congress could expand tax incentives included in the energy bill as a way to encourage growth of the refining industry

So that's another idea: The industry is waiting for a  government handout.

Another view:

Today investors are in no mood for refinery building even if funding were available," Arjun Murti, managing director of Goldman, Sachs Co., told a recent congressional hearing on the dearth of U.S. refining capacity. Any executive who might pursue a new $3 billion refinery risks his company's stocks taking a hit, said Murti.
The refinery industry's desire to increase profit margins is at the bottom of this flaccid growth in capacity over the last 30 years.  

According to Wyden's newer 2004 Report: No new refineries have been built in this country since 1976. The number of refineries dropped from 282 to 153 between 1977 and 2002. Refining capacity has increased 2.4% [by expanding existing refineries] while gasoline demand has lunged ahead 27%, over this time period. Internal documents from Chevron and Texaco in the mid-1990's said that they needed to decrease gasoline supplies and refining capacity in order to increase profit margins.

Whether you examine the BP data as Stuart has, or the EIA data (Wyden's numbers are from the EIA, considering the US only), the general glut in refinining capacity during the late 70's and 80's is evident.  it was this spare capacity that the major refiners wished to clamp down on, purely for profit considerations.  

From No End To Oil-Refining Bottlenecks

Dubai (heavy, sour) and WTI (light, sweet)

Here's his observation.
In these circumstances [the price differential], many refiners were reluctant to lift. While such heavy crude could be processed in the surplus primary distillation capacity around the world, the extra transport costs and the problem of disposing of the unwanted fuel oil only made such an exercise worthwhile at the right crude prices.

This development reflects an important fact in the oil market with significant price implications:
  • OPEC tends to produce heavy sour crudes
  • The cartel has a strong, vested interest in ensuring that the price of the heavier crudes do not fall too far below light crudes since it directly affects its members' revenue streams.
  • Tight product markets are likely to continue as long as OPEC seeks to protect the price of heavy sour crudes.
Now, as CalculatedRisk argued where we have this situation--

Raw Material --->> bottleneck --->> Finished Good

raising the production of raw material (heavy, sour crude) in this case could only lower its price because demand is constant--it is set by the bottleneck. The only way for heavy, sour prices to go up is to up the refinery capacity for these grades of oil and loosen the bottleneck.

Which is why Saudi Arabia is now in the refining business. Interestingly, these investments are in Asia. From OPEC tries to compete with Russia over China's oil needs:
Sheik Ahmad Fahd al-Ahmad al-Sabah, Kuwait's energy and oil minister and the president of the Organization of the Petroleum Exporting Countries, is scheduled to lead the group's first talks with China on Thursday as members like Saudi Arabia and Kuwait plan investments in Chinese refinery projects worth more than $8 billion. Saudi Arabia has used oil refinery investments to ensure sales in Japan and South Korea....

Saudi Aramco, the world's largest oil company by production, agreed in 2001 to expand a refinery in Fujian Province jointly owned with China Petroleum & Chemical, or Sinopec, and Exxon Mobil at a cost of $3.5 billion....

Kuwait and China have agreed to develop a refinery complex near Guangzhou, with the capacity to produce 200,000 to 400,000 barrels a day of gasoline and other fuels, Ahmad, the OPEC president, said in Kuwait on Dec. 12. That project, which would use Kuwait oil supplies, may cost as much as $5 billion, he said....
Saudi Arabia is also ramping up to build refining capacity to process their own output. (Sorry, no link on this). When the beloved and esteemed Ali Al-Naimi blames price hikes on refinery capacity, we can translate that as "we want to get as much money for our heavy, sour crude as you other guys are getting for that good light sweet stuff".

As for the US, from Exxon, BP Rebuff Pleas to Boost Exploration as Oil Demand Soars
A 200,000-barrel-a-day refinery would cost more than $2.5 billion and would require margins of about $9 a barrel for 20 years to provide an 8 percent to 10 percent return, says Joel Maness, 55, senior vice president of refining and supply at Sunoco Inc., the biggest refiner in the U.S. Northeast. U.S. refining margins have never averaged more than $9 a barrel for more than eight consecutive months.

``It would appear there's a purposeful effort to keep refining capacity tight because it increases profits,'' Senator Dianne Feinstein, a Democrat from California, told executives at the November hearing.

Mulva, whose ConocoPhillips controls 3 percent of global refining capacity, disputed the senator's contention. ``Not only are we adding capacity but we're modernizing our refineries so we can make more gasoline, jet fuel and heating oil,'' Mulva told Feinstein during the hearing.
I have no way of evaluating any of these statements but confusion reigns. Gratuitous Photo Opportunity.

Deep in thought...
Just a few quick notes (I'm mostly engaged in $$ producing work today), mostly in response to Jack's various remarks.

  • I agree about the changes in refineries, complexity, etc.  I agree there's been a lot of investment in debottlenecking and improving the output mix - especially in the US/Europe where we've moved away from using resid and fuel oil for heating, industrial processes etc, and are increasingly mainly using oil for transportation.
  • In my experience of raising money, investors don't usually say why they aren't investing, they just tell you to keep them posted and then stop returning your calls.  They play their cards close to their chest.
  • Investors all talk to each other all the time, do extensive due diligence before investing, and are familiar with the range of viewpoints on the market by the time they put money into a project.
  • Savvy members of the investment community like T. Boone Pickens, Matt Simmons, etc have been talking about peak oil for several years, as have energy analysts like Henry Groppe and Charles Maxwell.  I'd be very surprised if a sophisticated energy investing professional has not been at least aware of the hypothesis for some time.  They may or not buy into it, but it's bound to affect the FUD (Fear Uncertainty and Doubt) in the deal.
  • Forget all about explicit peak oil theories, and just look at the price signals.  The market signal for "we have a big problem with refinery capacity: build more" would be at least one refinery input would be cheap and at least one refinery output would be expensive.  That would make a screamingly obvious business case to build refinery capacity to improve the path from the cheap input to the expensive output.  We do not have that market signal even today - instead refineries face high but volatile output prices, and high but volatile input prices on all inputs.  Heavy crude is not cheap; even though the spread from light sweet has increased, it's still gone up a lot and it's still volatile (not what you want to see as an investor)  That's a market signal saying "wait and see" which is exactly what refiners have been doing in the US and Europe (by and large) in terms of increasing total capacity.
  • The places where building refinery capacity makes obvious sense even in a post peak scenario (with modest decline rates) are BRIC countries (stronger growth will be more resilient than that of US/Europe, so will be able to buy more of the world's supply over time), and the Middle East (move more of the value chain home).  That's exactly where new capacity is going, to a first order approximation.

In my mind, one is left with only two scenarios that make sense of the price and capacity history: 1) OPEC is getting better and more disciplined at price support on crude and decided to raise the price while claiming they weren't.  2) OPEC can't readily pump much more because they it's getting harder to overcome depletion in their fields, and no-one else can increase production quickly enough either.

I don't find 1) very plausible - what has changed to suddenly make OPEC more disciplined?  2) says imminent peak oil.

Awesome work again, Stuart, the results of which will assist us greatly in looking at the situation develop over the next five years.  Best work i've seen on capacity utilization.

However, u must admit that the data ends for the most part in Dec 2004.  While u would like to claim that capacity was not an issue (wrt supply) in latter 2005, we will not be able to see it 'til the new data is out.  But really, we know already what it will show.  We hit the 100% barrier.

Your studies this week by extrapolation illustrates that supply was stymied by capacity/production restrictions in the GoM and reduced exports from Iraq.  It is not a case of looking at the gloas half-full instead of half-empty.  There was a squeeze play this summer and i have no doubt that the growth rate in extraction will resume at a 1.25-mbd rate over the next five years.  Your work will also help us look at price movement as we go thru this period.  But capacity utilization is only one component.  OECD stocks are important as well.  And we have been mired in the 52-54 day reserve channel for much too long.  J-I-T (just-in-time) does not work well in the oil sector!

My Oil Depletion Scenario modelers continued their trend in 2005.  The optimists came down and the peakists revised up.  I expect this convergance of Outlooks to continue thru the decade.  I will start to worry when the Campbell/Laherrere/Kopplaar start to revise down to meet the TOD negativism.  They will be the canary in the mine ... not price ... not seasonal extraction downturns.    

My Oil Depletion Scenario modelers continued their trend in 2005.  The optimists came down and the peakists revised up.  I expect this convergence of Outlooks to continue thru the decade.

For Laherrere (ASPO moderate) and CERA, both using the same industry data source, the convergence may well be already there.

Laherrere (writing in 2003):

"If the present economic recession is protracted, demand could stay level for the next ten years, giving a bumpy production plateau of around 80 Mb/d. If supply is constrained by demand in this way, then the decline need not start before 2020."

CERA (writing in December 2005):

"The question of a worldwide peak in oil production continues to stimulate debate. Our outlook shows no evidence of a peak in worldwide oil production before 2020."


I agree with you that refinery capacity was likely signifantly tighter during 2005, especially with the hurricane outage.  My point is not that refinery capacity wasn't an issue, it's that it needn't have been an issue if efforts had been made to increase it that are entirely unremarkable by historical standards.  They weren't made because it didn't (and still doesn't) look like a good business proposition to expand refinery capacity aggressively in the US/Europe because input prices are too high and volatile (and output prices are affecting demand).  Thus the ROI on the deployed capital is not good enough to justify more effort than has been made (Dave has some nice quotes on this in one of his comments).  That fact has to be explained by restrictions in crude supply, not restrictions in refinery capacity.

On this we agree.  Just as nat'l gas producers sat on their hands after they saw three buck prices and over supply prevail, refiners may be looking for better margins.  A few weeeks ago i suggested that we are going to see a growing disparity between crude price and retail prices.  Crude falling due to global refinery saturation.  And line-ups at the refinery gates.  Your work underscores that scenario unfolding.
Freddy has a good point. If Colin Campbell has the liquids peak at 2010, and Laherrere at 2015, what's the motivation for analysis like the OP which assumes that we are at (or almost at) the peak? Is Stuart 10 years more pessimistic than Laherrere? And if so, why?

To the lay reader, there is a huge emotional difference between "we're probably peaking now" and "we'll probably peak in 2015". In the latter case, you might quit worrying about it, and check back in on the blogs around 2013.

Re: "Forget all about explicit peak oil theories, and just look at the price signals...."

OK, I done that.

Re: "1) OPEC is getting better and more disciplined at price support on crude and decided to raise the price while claiming they weren't. 2) OPEC can't readily pump much more because they it's getting harder to overcome depletion in their fields, and no-one else can increase production quickly enough either."

Yet as I pointed out, OPEC is getting more heavily into the refining business. Why would that be if your #2 scenario is correct? They're making some heavy investments so they have heard "the market signal for 'we have a big problem with refinery capacity: build more'". Especially in Asia which is where most of the demand growth is. The beloved & esteemed Ali Al-Naimi is, after all is said and done, a businessman. He can't do business selling heavy crude to the US or Europe so he has decided to do business with China where ironically he has more leverage. From The Emirates Economist
The Organization of Petroleum Exporting Countries, including Abu Dhabi and the rest of the United Arab Emirates, is considering plans to build at least 10 new refineries, seeking to relieve strains on existing plants to meet fuel demand. U.S. Federal Reserve Chairman Alan Greenspan yesterday said world refinery capacity is "worrisome."

The OPEC refineries, if built, would increase global processing capacity by 2.4 million barrels a day, or 2.8 percent, by 2011. World crude oil prices have doubled in the last two years....

"This is a good time to be in the refinery business as margins are high and will remain high for the next three to four years," [Mohammed Al ] Khaily [managing director of The Persian Gulf emirate's International Petroleum Investment Co.] said in Abu Dhabi, the largest member of the United Arab Emirates federation, on Oct. 17 [2005].
After you're done making some $$, could you explain your view of this to me? Why would the US/Europe and OPEC be so wildly at variance on this issue? Please, if anyone else has some thoughts on this, please speak up.
I alluded to this briefly in my earlier comment, but again.  Suppose there's a peak in the next year or two.  Suppose decline is moderate afterwards, and the world starts vigorously adapting (forget the SUV honey, we need a Prius etc).  The US and Europe can almost certainly actually reduce their oil consumption while maintaining some economic growth, as long as the required rate of improvement in efficiency is not more than 5% a year, give or take.  That leaves more oil for China and India if the overall production decline rate, on average, is low (significantly less than 5%) in the near term (ie if you believe what the Hubbert linearization predicts).  The China/India growth, and the process of globalization will not be stopped by a low-post-peak decline rate (IMO).  The same reasons why it's cheaper to make goods in China and perform services in India now will still be true post peak.  So their oil use will continue to increase, even as the US/Europe decrease oil use via increased efficiency (plus we have a big debt-fueled housing bubble to burst which will probably crimp our style for quite a while).  So China/India will need more oil refinery capacity, and the US/Europe will need less.

As far as the Saudis - how can they lose?  When the world is very thirsty because there isn't quite enough oil and they have a big chunk of the remaining resource (even if it's not as big a chunk as they claim, it's still big), they stand to make the most profit by selling us refined products instead of crude and gouging us on the refinery margin as well as on the crude itself.  There's no reason they have to send us any crude - they can just send us products, wait for our refineries to go out of business, and then clean up.  That's what I'd do if I were a ruthless business-person in charge of Saudi oil.

Well, thanks for your response. But...

You seem to place a lot of faith in
The US and Europe can almost certainly actually reduce their oil consumption while maintaining some economic growth, as long as the required rate of improvement in efficiency is not more than 5% a year, give or take.
I am familiar with your view here. If you are right about this, then US/European refining capacity will be adequate and it will be OPEC--who are increasing their own refining capacity--meeting more and more of our refined products demand. This has the obvious consequence of making us more dependent on them than ever before. On the other hand, if the US expands its own refining capacity, it is more capable of importing and processing oil from a larger, more diverse set of sources (like West Africa). This would seem to be a desirable goal that makes us much more secure--it is axiomatic not to put all your eggs in one basket.

So, IMO, the picure you paint points toward a shit-storm crisis of huge proportions as we go forward--especially in light of the geopolitical volatility of the Middle East. Now surely some of the big financial investors in the Western World might have the foresight to acknowledge and take steps to mitigate this problem?

In your view, apparently not.
I should add, to make my comment more clear, that I do not agree with your assumptions about the elasticity of demand for oil in the US. I certainly can not envision much GDP growth (even at smaller rates) commensurate with small (3 to 5%) decreases in our oil consumption. Recession is the far more likely outcome. Our liquids demand will go up year-on year in any case, as it has in recent years, albeit more slowly than the increased consumption rate in Asia. Therefore, we get the dangerous dependency on greater imports of refined products from OPEC. The assumption that greater efficiency in usage will save the day--for what period of time?--is misguided in my view despite the modelling you have done of these issues.

So, given the security issues I spoke of above and the view that US demand is fairly inelastic, I would say that the shit-storm I spoke of has a middle-range probability in the next few years in light of the reluctance and inability of the US to expand refinery capacity and increase demand elasticity.
The traditional assertion is that demand for oil is inelastic in the short-term and elastic in the long-term. If this is accurate, and I think it is, the questions become:

  1. Are people already getting adequate price signals to change behavior?
  2. How much time is left before supply constraints force the small (3-5%) consumption drop?

I think that 1) is starting to happen. As long as the time frame for 2) is 5-10 years, I do see the ability to adjust.

The reason I think this is because the US has a huge amount of energy consumption that is basically a luxury. It could be shaved off without disruptive economic impacts (on the economy as a whole). If oil prices stay high, these prices willgradually be passed on all products that depend on oil including personal trabsportation, home heating, and purchase of goods that have high energy inputs (in production and/or transportation).

I am not saying this will be magic, easy or guaranteed - or that some people won't get hurt. I do think that the US economy can grow despite diminishing oil supplies if the adjustment period is smooth.


In your blow-up of production growth in the last 3 years, growth is negative by the end of 2005.  I find this interesting. We don't see any sustained increase in demand over the past few years, another reason why refinery majors would be VERY reluctant to increase capacity, to ensure their profit margins stay high (as they have done for 20 years) while capacity remains tight .  Peak oil is not necessary to explain this because it's been going on for decades.  

i need to correct myself on this.  i mistakenly thought the Y-on-Y graph represented demand while it is actually supply.  my point about low demand growth isn't correct (demand is still increasing, while supply growth is stagnant recently), but the fact remains that refiners are in an ever better position for super-high profits, as long as they keep supply growth low.  
Regarding that argument made by Calculated Risk and
cited by Stuart: is there some reason why cartel
behavior by the petroleum industry as a whole is
inconceivable, as opposed to just refiners?  If not,
a peak (inevitable sometime for purely geologic reasons) could be engineered to occur at the time and
in the manner most profitable for the industry and most
politically oppertune for governments with influence
over the industry.  Is that sort of coordination
No offense intended but your comment here is what worries me about Stuart's story and some other remarks made here.

Doesn't all this imply some, if not a conspiracy, but rather a tacit agreement among all those in a position to invest in refinery infrastructure (at least in the West) that peak oil risks do not justify such investments? As Stuart says, all these investors tend to talk to each other and are obviously aware of the forecasts made by Simmons, Groppe and others. On the other hand, there are optimistic assertions made by CERA, the USGS, Lynch, IEA et. al. There's surely a wide diversity of views on this among the big investors.

I for one simply don't believe that peak oil is taken seriously enough by everyone in the world of big finance to make these kinds of claims.
In answer to Stuart's closing question about why oil consumption growth slowed down last year, in the simplest form I don't see much of a mystery. Oil reached a very high price, which naturally and understandably reduced (or, as some people here like to say, "destroyed") demand. This is exactly what standard economics would predict, but it is not the interesting part of the question.

The real question is why did oil prices get so high this past year, and in general why have they been on such a steady upward trend since hitting a low of about $10 (in today's money) in the late 90s? Well, I don't know! I have certainly heard a number of opinions about it, but IMO most explanations offered for financial trends are not very convincing and credible.

However I would suggest that the price rise this past year should be seen in the context of the steady price rises in previous years. In other words, it is the continuation of a trend, not something new. So it seems logical to me that the cause of this past year's high prices is probably closely related to the forces which have caused the price to rise over the previous five year period.

Now, if this cause is in fact as Stuart suggests "peak oil", a physical limit on how much oil can be produced, then that means that the first stirrings of peak oil were felt five years ago. I'm not sure that is a plausible explanation though. Few people invoke peak oil to explain the early stages of the climb from $10 oil. We could argue that the beginning of the price rise had one cause and only the recent increases were due to peak oil, but Occam's Razor somewhat discourages such multiplication.

The alternative explanation I have seen is that the low prices of the 90s, after the price collapses of the 80s, made the industry risk-averse and so there was under-investment in development projects in the 90s. Since there is such a long lead time between the initiation of a project and when it reaches full production, the result was an "oil drought" which we are still experiencing today. This explanation predicts that recent high prices will motivate a surge of new projects and that there will be a glut of oil in the near future.

That's the great thing about explanations, there are always so many of them that we are free to choose the one that best fits our predispositions and prejudices.

As a believer in efficient markets, wouldn't you expect futures market arbitrage to kick in some years ahead of time?
Yes, I would expect the markets to be forward-looking. But most market participants are reasonably open about what is motivating their investments and AFAIK there were virtually no investors saying that they were worried about peak oil back then. Whereas today I think there is a growing sense of concern about long term oil supplies among traders, which is seen as a factor working to drive up prices.

It's conceivable that several years ago, there was somewhat "secret" and closely held knowledge about peak oil among insiders who were taking positions in the market in advance of greater public awareness of the issue. Then you might see something like this, as their investments, small at first, gradually began to make an impact. Prices would begin to rise even without widespread understanding of what is driving them. Then eventually people would clue in to what is happening and the effects would grow.

As I say, it is a conceivable story but I'm not sure it is a persuasive one. It requires something of a conspiracy among "insiders" and a degree of successful secrecy that is not often found in the world. It's also difficult to come up with a specific scenario for the time frame over which this happened. I'm not sure there is any historical precedent for an investment conspiracy based on private knowledge that managed to keep its advantageous secret over a period of years.

I have a different, but equally subjective, impression.  My sense is that mentions of peak oil in the business and mainstream media have been building for several years.  We need a way to assess the frequency of peak oil mentions more quantitatively (the kind of thing that Robert Shiller does in tracing the level of awareness of markets about different points).  One can do it in services like LexisNexis, but I don't know a ready way to do it for free. only goes back six months.
God Damn!

Considering Halfin's comment and this response, I feel like I'm entering an alternative universe. Halfin says, responsibly in my view, that a "conspiracy" is improbable. Thank God.

Re: "I have a different, but equally subjective, impression. My sense is that mentions of peak oil in the business and mainstream media have been building for several years".

Of course. This is because prices are rising rapidly year-on-year and it's become obvious that supply is barely keeping up with demand. That's it, that's the whole ball of wax. That's why there's more publicity. But to even talk about a conspiracy and make a serious reply to that--even if we are dismissing it--makes the whole serious discussion about Peak Oil vulnerable to all sorts of crazies that have latched on to the idea. I'm sorry, but I'd have to say, Stuart, that your post opened this Pandora's Box and I'm trying very hard to close it.
I'm not sure what's bothering you Dave. Halfin was arguing that peak oil could not have been a factor in oil prices in the 2002-2004 timeframe (more-or-less) because no-one was talking about it. That's his subjective impression (I assume - he cites no source). I have a different subjective impression - people were talking about it, and therefore it could have been affecting the behavior of, say, energy hedge funds, and thus prices. Then I'm pointing out that my subjective impression versus Halfin's is not much of a basis for deciding the question, and wishing we had better data to assess the issue (which I mentioned in the hope that somebody else might know of some).
A search on ABI Inform database. ABI/INFORM Global indexes and abstracts over 1,400 business and economics journals, magazines, and trade publications. In addition, it provides the full text and/or page images of the articles from about half of these periodicals.

Baseline: Number of articles mentioning the word "oil" in citation or article text calendar year 2005: 31,642

Number of articles mentioning either "peak oil" or "oil depletion" in citation or article text for calendar year:

2005        74  (out of 31,642 mentioning "oil")
2004        39
2003        24
2002        15
2001        19
2000        14
1999        12
1998        10
1997        12
1996        9
1995        7
1994        15
1993        6
1992        8
1991        1
1990        0

So mentions of peak oil are up dramatically--but still trivially small in the absolute, when examining business, academic, and trade journals. I'm sure the numbers would be much higher in mainstream media, but probably with similar trends.

Excellent! Thanks for gathering that.
Halfin: If you go back and look at Matt Simmons' speeches you can find him making very clear statements on the issues to industry and business groups all over the world from at least 1999 on. He certainly talks at length about the crappy returns they'd all gotten in the 80s and 90s. But he also talks at length about depletion and the possibility that the world might not be far from peak oil production, and North America was going to deplete in NG. For example, read this address to the Offshore Technology Conference in 2001 (which is probably the biggest oil industry trade show). So this message has been out there in industry all through the relevant timeframe. Not saying everyone believed him, or does now, but a picture of that belief slowly diffusing through the energy markets and slowly increasing prices seems plausible to me.
In fact, thinking about, there's really a complete positive feedback loop here. We know (eg from Oil Drum web stats) that high prices stimulate interest in peak oil. But it seems obvious that belief in peak oil by market participants will tend to drive up prices, pushing more of them to study peak oil, perhaps believe it, and drive up prices further.

Of course if belief in a near-term peak was false, what would happen is that suppliers would be delighted and jump in to try and capture market share while the prices were so excellent. This negative feedback loop on prices would depress them again after a while. It is the fact that they have not been able to do so to good effect that leaves the hypothesis intact.

Correct. We are just getting to market stabilisation now (at the $62 to $64 range) after the 'tweaking' in response to Katrina. The next few weeks will be very interesting: does the market go down or does it try to go up (and, if so, are there signs of manipulation to force it down). PO is still mostly a fringe concept in 'market' terms, God help us when the markets begin to think otherwise.

$70 oil in March looks a sound bet.

Jim Hamilton at Econbrowser has a good piece discussing the fact that oil company investment is not one would expect under the circumstances here
How about this:
1 Prices collapsed in 1998 because OPEC squabbled and over-produced.
2 Prices recovered in 2000 because OPEC settled the squabble and reduced production.
3 Prices declined in 2001 because of the crash.
4 The surprising new demand from China/India, not expected by any market participants, let to ever higher demand and prices through mid-2005 as OPEC struggled to meet demand, ultimately producing all they could, save only some over-priced heavy sour from Saudi.
5 K/R raised prices further, to the extent that US gasoline demand declined for the first time since around 1980.
6 Reduced demand from high prices lead to rising stocks of both crude and product, lowering prices. Lower prices then lead to an immediate increase in demand, but crude stocks continued to rise.
7 Rising stocks is a red flag to OPEC, and reminded them of their raison d'etre. They are particularly alarmed because prices were falling in winter, even before the onset of the spring shoulder season. Accordingly, and confident that $60 poses no harm to their more important customers, they seem likely to reduce output following their end-Jan meeting, no doubt pleasing peak oilers as well as big and small oil alike.

None of this, by itself, means peak oil is, or is not, around the corner. On balance, the evidence seems to indicate it may be coming soon, but I can understand fears of oil CEO's that investments today, at high prices, will lose them money later. Executives of mid and small-sized companies, in particular, no coubt remember how they, or others, lost their jobs in the 1998 crash. Nearly all of these executives are nearing retirement - no time to risk all on new ventures.

Sales of existing properties by a large company to a smaller one means nothing by itself - this is traditional; as properties require ever higher amounts of labor to develop the last dregs, they are routinely passed to those with lower costs.  There is no doubt that costs are rising - perhaps the large companies just need time to convince themselves the high prices are here to stay.  OPEC might help in this matter. Meanwhile, buying back stock keeps stock holders happy and usefully boosts executive options.

I'm glad you bought up the Chinese/India demand surprise thing. This is another thing I hear repeated over and over again in the MSM. Can someone please point out to me which part of these curves was the part where the unexpected surprise happened?

Here's the YoY growth rates. Definitely China had a big 04, but not that much out of their range of variability, and India has had a rather lower growth the last few years.

Overall world demand growth was 4.2% year on year in 2004 - certainly strong, but by no means off the scale of recent historical exerience.

Maybe growing Chinese demand had a sudden impact when they stopped exporting and began importing oil, and the imports rose sharply. OPEC saw a significant competitor turn into a growing customer.
And, the larger growth rates earlier, during the seventies, came on a very small base. The affect China's growing economy is having, whether on resource imports or competition, affects the rest of the world much more now for the same reason. 10%/year growth was not much in the eighties, but after a couple of decades go by, the elephant is no longer a baby. Imagine another fifty years... no worries, we'll all be out of oil by then.
"I can think of only two reasons why we might not have chosen to increase refinery capacity in the way that we obviously could. One is a secret refining cartel. The other is peak oil."

Really? You couldn't think of any other explanation?

The economists will argue that obviously there is a fear on the part of the refining industry that market prices for refined oil products will collapse again, as they did before; that fear is the obvious explanation for the reluctance to invest in new refineries. It's the exact opposite of a fear of Peak Oil.

I don't necessarily subscibe to that argument. But it is the obvious argument, and to ignore it by saying it didn't occur to you is not, I believe, a good way to engage Peak Oil doubters into a meaningful discussion. Rather, it will make it seem as if your conclusions are formulated upon heavily biased premises.

Actually, what mainstream neoclassical economists tend to point out is that a refiner afraid of that risk can hedge against the risk in the futures market. They prefer to model people as rational agents who do not act out of emotions like fear. However, a behavioral economist might well agree with your argument, and I think it has some plausability. What I would ask is this, though: surely in the late nineties, the enormous overcapacity of the industry in the early eighties would have been much clearer. And yet refinery capacity increased faster then than it has been lately. This story from last summer that Dave quoted speaks volumes:
"The 10-year average return on investment in the (refining) industry is about 5.5 percent, about what investors could receive by investing in government bonds with little or no risk," says Bob Slaughter president of the National Petrochemical and Refiners Association. But some independent refiners, such as Valero Energy Corp., have been able to reap considerably higher rates of return _ buying old refineries at bargain prices, as little as 20 cents on the dollar. Valero, the country's largest independent refiner based in San Antonio, Texas, has grown rapidly since 1997 by acquiring undervalued refineries, making environmental improvements and expanding capacity. Such was the case with its purchase of the Orion refinery in Louisiana. "We paid far less (for it) than the replacement cost. If we had paid full replacement cost for it, it would not be doing well at all," said Gene Edwards, a Valero senior vice president. Would Valero ever build a new refinery? "I don't see anyone building a refinery in the U.S. _ maybe overseas," said Edwards in an interview.
(emphasis mine).

There is no enthusiasm for increasing capacity. What would it take to make them enthused? Higher ROI. What would make the ROI higher? A larger spread between crude and products. In short, the lack of refinery capacity growth is an effect of the high cost of crude (at least heavy grades), not a cause of it.

Fascinating insight Stuart!
I dunno. As you pointed out separately, the spread between heavy and light grades has been increasing, also explaining why Valero is making so much money expanding capacity. I agree the high cost of sweet/light vs. product discourages expansion of refineries that handle the premium grades. As far as I can tell, teh new refineries planned around the world are being designed for the cheaper grades, also acklnowledging that peak s/l is behind us.
Do you have the data to see what is happening to total product production as the world's crude moves toward heavy/sour?  If peak crude output to date is May 05, isn't it likely that product volumes peak to date is sometime earlier? Product flows switch around based on price, so maybe gasoline/distillate combination?
To answer Stuart: yeah the spreads are roughly the saame between heavy and light but another good indicator is the releases from the spr that weren't takers for the heavier crudes.

Yes, there will always be a balance - as far as refiners see it - with capacity.

Most new refineries are being built in the m/east but also in Indonesia, China (with aramco, exxon and others), Nigeria. In m/e Qatar, Saudi, Kuwait, UAE all expanding, and more countries too. Iran desperately needs refining capacity - it impoirts $7bn a year of refined product. Also many of these nations have a rising young, population.

But big new expansions are online in existing US refineries, Valero announced one recently if i remember...

As for the media - its very different to just 3 or 4 years regards oil and energy reporting etc...but it could be much better...