This Week in Petroleum 5-2-07

12 Weekly Declines in a Row: Not a Pretty Picture

Update: My Limb Has Cracked

Two weeks ago, I went out on a limb and said that gasoline inventories would turn up within two weeks. Last week saw a pretty steep 2.8 million barrel draw. I stuck to my prediction for this week, and while the draw did drop to 1.1 million barrels (a month ago it was in the 5 million barrel range) it was a draw nonetheless. So, my limb broke. But there was a reason I included the section “Why I Might Be Wrong.” There is a lot of art involved in this forecasting, and many factors to consider. Otherwise, the analysts would always get it right, when in reality we have seen them miss badly on a regular basis.

I do think we are pretty close to the turning point, because the rate of the inventory draw has been dramatically slowing. However, the fact that we had an inventory draw this week continues the unprecedented situation we find ourselves in. On an absolute volume basis, in the past 20 years we have only had two readings in the spring that were lower than this. There were a couple of weeks in the first half of April 2001 that showed slightly lower gasoline inventories. But demand has increased substantially since then. On a days of supply basis, this week’s inventory number is the lowest we have ever seen in the spring, prior to peak summer driving season.

There are two primary reasons for this. One, is that refinery utilization is running a little behind normal. So, supply is down just a bit. But, demand is up. Higher demand + lower supply equals the situation we now find ourselves in. And while demand has been slowing, it is still running ahead of last year’s demand figures. That can only mean one thing: Higher gasoline prices until demand is brought back in line with available supplies.

Here were the highlights of the text report:

U.S. commercial crude oil inventories (excluding those in the Strategic Petroleum Reserve) rose by 1.1 million barrels compared to the previous week. At 335.6 million barrels, U.S. crude oil inventories are in the upper half of the average range for this time of year. Total motor gasoline inventories fell by 1.1 million barrels last week, and remain well below the lower end of the average range. Distillate fuel inventories inched lower by 0.2 million barrels per day, and are just below the upper end of the average range for this time of year.

So, there’s the inventory situation. But, crude imports continue to run higher than normal:

U.S. crude oil imports averaged nearly 10.3 million barrels per day last week, up 229,000 barrels per day from the previous week. Over the last four weeks, crude oil imports have averaged 10.0 million barrels per day, or 205,000 barrels per day more than averaged over the same four-week period last year.

And total products supplied were up, just not enough to meet demand:

Total products supplied over the last four-week period has averaged nearly 20.9 million barrels per day, or 3.3 percent above the same period last year. Over the last four weeks, motor gasoline demand has averaged nearly 9.3 million barrels per day, or 1.6 percent above the same period last year. Distillate fuel demand has averaged over 4.3 million barrels per day over the last four weeks, up 5.4 percent compared to the same period last year.

You can see Doug MacIntyre’s take on these numbers in This Week in Petroleum, which was published just a few moments ago. Not enough focus on the gasoline inventories, in my opinion, but a good discussion of benchmarks.


As the gravity of the gasoline inventory situation sinks into the public consciousness, more people than ever before will be closely watching this week’s inventory report. More and more stories are appearing in the media highlighting what I have been saying for a while now: Gasoline inventories are at unprecedented levels for this time of year. Two weeks ago I went out on a limb and said that inventories would turn upward within 2 weeks. While they did fall again last week, I am sticking to my prediction that they will turn back upward this week.

Why I am Right

A number of factors have combined to put us in the current inventory situation. Probably the most significant is that demand throughout the winter was at all-time highs. While we did build a nice inventory cushion going into spring turnaround season, it depleted much faster than was typical once turnaround season actually started. It was the rate of the inventory drawdown that has led me to speculate that prices would be headed much higher. And so far, the price data support this speculation.

However, I can see some trends starting to turn. As prices have spiked, gasoline demand has started to slow. Demand has fallen each week since the end of March, and is now 300,000 barrels a day lower than it was a month ago. While last week’s demand data is still running slightly ahead of the data from a year ago, that gap has closed. So demand is now more in line with where it was last year.

The second major factor putting us in the current position is that refinery utilization has been lower than normal (due to “ineptness” and “greed”, according to the clueless bunch at the FTCR; I wonder why no oil company has hired them as a consultant since they seem to know how to run a refinery). After climbing above the 90% utilization mark 2 weeks ago, a couple of refinery outages dropped that number back below 88% last week. Actually looking back to last year, we saw similar numbers, but then in May and into June utilization climbed back above 90%. We should see a similar pattern emerge this year, barring something unforeseen at one or two major refineries.

The third item to consider is imports. First quarter gasoline imports ran well behind last year’s numbers. That has helped put us in this situation. Starting in May of 2006, imports spiked up (eventually by half a million barrels a day) and eased the supply constraints. Looking back, however, last year’s import situation was an anomaly. Yet we are going to need this anomaly to repeat itself this year if supply and demand are to be brought into balance without prices having to go up another $0.50 a gallon. At current prices, though, I expect imports to start hitting the shores to take advantage.

I will note that it was last year’s April 28th report that showed a gain in gasoline inventories after they had fallen for seven straight weeks. Falling gasoline inventories at this time of year are typical because of turnaround season and the spring vapor pressure turnover. High demand has changed that picture this year and caused a steeper fall than normal. But this week I think we will see that gasoline stocks gain, reversing the trend just as they did during this week in 2006.

Why I Might Be Wrong

I think that demand for last week is likely to come in close to the level of a year ago (especially with the price spike following last week’s report), but the import and refinery utilization pictures are less clear. The import situation is further clouded by the threat of a refinery strike in Belgium. (Belgium, while a small country, does refine a large amount of oil). Refiners there may decide that they can’t afford to export gasoline to the U.S. with this hanging over their heads. Furthermore, a refinery in Oklahoma did go down last week as a result of a massive fire. While this is a very small refinery (50,000 bpd) every little bit counts when the market is as tight as it is.

I will note that according to a survey of analysts by Bloomberg:

U.S. gasoline stockpiles probably fell for a 12th week, as rising demand and breakdowns hampered refiners’ efforts to store fuel. Demand in the four weeks ended April 20 averaged 9.3 million barrels a day, 2.3 percent higher than a year earlier, the department said last week.

Gasoline supplies probably declined by 1.3 million barrels, based on the median estimate among the 13 analysts surveyed. Inventories dropped 2.79 million barrels to 194.2 million barrels the week before, leaving them 7.2 percent below the five-year average.

They do, however, forecast that crude inventories will continue to build:

Oil inventories in the U.S., the world’s biggest consumer, rose to a five-month high last week. An Energy Department report tomorrow will probably show the above-average oil inventories gained an extra 1.5 million barrels last week, according to a Bloomberg News survey of analysts.

Of course this is inversely related to their prediction of a gasoline inventory decline. When refinery utilization is high, they tend to draw down crude inventories and build gasoline inventories. So it would appear – reading between the lines – that the analysts believe that refinery utilization will come in low again this week.

Is This Evidence of Peak Oil?

In a word, no. I have seen a lot of people make this insinuation – that the draw down in gasoline inventories supports the contention that oil production worldwide has peaked. However, this is a real misreading of the situation. First of all, oil inventories are quite healthy, well above the average for this time of year. Crude oil imports are running slightly ahead of their values of a year ago. China also recently announced that they are currently importing record volumes of oil.

“OK”, you say, “But how do I know that crude quality hasn’t deteriorated and therefore gasoline yields are lower? How do I know that we aren’t simply outbidding other countries for crude?”

Well, on the quality issue, we have the data. Crude has heavied up a little since 2000. But refiners have been installing cokers and hydrocrackers, which will increase the demand for heavy oil. And in the past 4 years or so, crude gravity has been relatively constant. So it isn’t that refiners are suddenly having to deal with undesirable heavy crudes that are affecting their gasoline yields. You may read some sporadic accounts of this, but they are the exception. Looking at the sulfur situation, we see a similar picture. Sulfur content has been relatively constant for several years.

So, what about gasoline imports? If certain countries have peaked, perhaps they are using more of their gasoline internally. No. If we look back at the past several years, gasoline imports for this time of year are quite normal. In fact, if you look back a few years, with the exception of last year gasoline imports are running well ahead of the average over the past 5 years. So that explanation doesn’t add up either.

The gasoline inventory situation just doesn’t provide any peak oil evidence one way or the other. If we were looking at the identical situation with crude inventories – that is to say inventories below the 5-year average with crude prices spiking higher and higher – that would be much more indicative of support for peak oil. What we have right now is more of a refining/record demand issue.

The Price Debate

I have been warning of much higher prices for weeks. Doug MacIntyre, author of This Week in Petroleum, has dropped by several times and we have engaged in some debate about prices heading into summer. Doug’s contention was that prices are likely to moderate going into summer. As Doug had written:

…recent history has indicated that we are not guaranteed of seeing higher prices on Memorial Day or around July 4, than we see in mid-April. I think retail prices will be heading down soon, possibly as early as our next price survey on Monday…

As I responded:

I would say that it just depends on inventories. We are going into high-demand season in worse shape than we have been in for a while. Unless the trend reverses (gasoline stocks have declined for 10 weeks now) then I do expect prices to continue trending higher. Although on a week to week basis we might see no change or even a decline in prices, I think on a month to month basis falling inventories will continue to keep upward pressure on prices until the demand is brought back into balance with available supply.

To which Doug wrote back:

We agree a lot more than we disagree. Inventories need to improve soon. Also, we need to be relatively free of any major refinery problems this summer. But I agree that inventories are critical in shaping the price outlook for this summer.

Since then, there have been 2 price surveys. The first was almost flat, with prices declining by less than a cent per gallon. So, Doug gets a very slight nod on that one. But this past week, the average U.S. gasoline price was up by over a dime a gallon. That is a steep rise for a single week. So, I get the nod on that one, and for the 2-week period gasoline prices are up by 9 cents a gallon. And if inventories drop again this week, we are definitely headed for record prices this summer. At this stage, we may be headed there regardless of what inventories do over the next month.

I will update this at some point after the report is released tomorrow, but I will be tied up immediately after the release of the report. (For the next 3 days, I will be learning how to climb out of a helicopter underwater). It will probably be 2 or 3 hours after the release before I do the update.

29 thoughts on “This Week in Petroleum 5-2-07”

  1. Robert,

    Me again. 2 weeks ago, refinery runs were above 15.4 and on the way up, or so I thought. Last week’s data showed refinery runs falling to just under 15.1. I did not expect such a decline and neither did the market, which is why prices shot up, both on the wholesale and retail levels. This week’s data on gasoline inventories will be important, but also important will be data on refinery runs and gasoline production. EIA is now expecting prices to go up at least the next week or two, at the minimum, and I wouldn’t be surprised to see prices go up every week up through Memorial Day. But the data this week will be critical, as we will only have 4 weeks of data left before Memorial Day, the start of the peak driving season here in the States.


    P.S. I’m gald you don’t seem to mind me chiming in every now and then, but a blog that keys off of TWIP every week got my attention and the quality of your analysis kept my attention!

  2. I’m gald you don’t seem to mind me chiming in every now and then, but a blog that keys off of TWIP every week got my attention and the quality of your analysis kept my attention!


    I am very glad that you drop by, and I know a number of readers are as well. I have seen your visits here mentioned on a number of other boards as well. Personally, This Week in Petroleum is one of the key pieces of information I look to each week to understand what’s happening, and it was also something my former refinery analyzed in detail each week as soon as the data came out.

    I am predicting an upturn in gasoline inventories tomorrow, but it looks like most of the analysts polled by Bloomberg think it’s going to be another decline.

    Cheers, Robert

  3. Fine analysis.

    If this is a refinery/demand issue, which seems likely, then what is being done to increase United States’ refinery capacity?

    In the future, will we be depending on foreign countries for imports? Or do you know of any refinery expansion projects currently underway?


  4. Or do you know of any refinery expansion projects currently underway?

    There are loads of expansion projects underway, and capacity has expanded by a lot over the years. See:

    Tyson Slocum is Wrong

    In the past 10 years, refining capacity has increased by an average of 200,000 barrels each year. The problem is that even this level of expansion has been unable to keep up with demand.

    On the import question, I think it is likely that we will be more dependent upon finished product inventories in the future. Saudi Arabia is evaluating some major refinery projects. So, instead of importing Saudi oil, we may be importing Saudi gasoline.

    Cheers, Robert

  5. Regarding higher gasoline demand at the end of March, being that Easter was 3/23 and man schools spring break was the week following, I wonder it that is why gasoline demand was highest for week ending 3/30.

    Is there historical evidence for a slight consumption bumb around spring break calendars?

  6. I wonder it that is why gasoline demand was highest for week ending 3/30.

    The thing is, there has been a weekly drop each week since then. It wasn’t just a one-time spike. It looks like it is correlated to price.

    Cheers, Robert

  7. This week’s report is out. Very nice job discussing the issues, but do you like ketchup with your hat? 🙂

  8. BenjaminCole
    Plenty of crude around, that’s why the price drop. Between 1979 and 1983 as a result of the price spike of the late 1970s there was an 11.2% (6.5 million b/d) decline in worldwide crude consumption. This price shock not as bad, but appears more sustained. And much more technology out there now. Bio-fuels coming on, and are mandated to make up 20 percent of Europe’s diesel consumption within 15 years. India doing the same thing. Even the USA might o something.
    Meanwhile, the worldwide marginal cost of lifting a barrel of oil is probably around $5, It was $3.57 a barrel back in 2003, according to EIA.
    There may be “peak oil” for the light sweet stuff, but even that is a maybe in the next 30 years. But there is gobs of heavy stuff.
    And do not forget demand. With plug-in hybrids on the horizon, the real story may be “peak demand” not “peak supply.”
    In fact, I think there may be a terrific argument that we have reached “peak demand” and i would love to see a msartie out there begin to flesh out this point.

  9. Your inventory graph does not show the most recent data point, probably because you grabbed last week’s version which was cached somewhere. I find that I have to refresh my browser every time I load TWIP in order to properly update the graphs.

    The last data point shows what could be the beginning of a bottoming-out process for gasoline inventories because the rate of inventory decline has markedly slowed this week.

  10. Guess I will tell my wife to fill up today ahead of the price increases that are coming in the next 24 hours.

    Not good news.

    While waiting for the report, I was trolling EIA and found this presentation:
    Are Refiners Entering a Golden Age or a Short Cycle

    I wanted to point out a couple of slides. Look at slide 6. This shows the relationship between excess OPEC capacity and WTI price. This shows how tight crude supplies are and the folly of the argument that “xyz crude area shouldn’t be explored or produced”. Adding 1 or 2 million barrels of non-OPEC production (think 1002 area in Alaska) would drive prices back down to $30/barrel and reduce some of the mischief coming out of the Middle East and South America.

    Slide 25 is for the folks at FTCR and Tyson Slocum. The industry IS planning to add capacity.

    I would promote opening the 1002 and other areas in the US for development along with some sort of oil import fee or carbon tax on transportation fuels and an increase in auto fuel efficiency standards. We need to keep prices high to encourage conservation and innovation but also need to reduce the amount of money funding terrorism and the spread of socialism.

  11. Robert,

    I can’t write every report on gasoline! We’re only in early May and there will be plenty of opportunities to discuss gasoline, unfortunately.


  12. There’s probably no easy answer to this without knowing the detailed profiles of US refineries and the crude feedstock input, but…

    Is there any guestimate around as to the theoretical maximum domestic US gasoline output if refineries were running at 100% utilization?

    Basically I’m wondering how much leeway refiners have to devote production to gasoline at the expense of other outputs.

  13. Robert

    I am glad to see that Doug has picked up on the Tapis spread in his weekly article.

    It is firmly stuck in that $74-$76 a barrel range.

    Doug if you are watching what is your reading on Tapis prices?

    Also where can I get refinery utilisation numbers for Asia?


  14. Is there any guestimate around as to the theoretical maximum domestic US gasoline output if refineries were running at 100% utilization?

    Actually, that’s probably a fairly easy exercise. At these prices, I can promise you that refiners are making all the gasoline they possibly can. So, take this week’s gasoline production number and divide it by the utilization number of 88.25%. By my calculations, that gives right at 10 million barrels per day of U.S. gasoline production capacity.

    It won’t be exact, but pretty close to a maximum current theoretical capacity. Of course, refiners can shift some product from diesel to gasoline, but not without causing shortages in the diesel market.

    Cheers, Robert

  15. It won’t be exact, but pretty close to a maximum current theoretical capacity.

    After thinking about that a bit more, if I recall correctly those production numbers do include imported blend stocks. I am not sure if that number is available so it could be subtracted out to get a real production capability.

    Actually, I guess we could figure it out. We have the production numbers, the import numbers, and how much inventories fell. I think from that we can figure out how much production then came from U.S. refineries. But it’s getting late here, so I will have to revisit that tomorrow.

    Cheers, Robert

  16. Two weeks ago I went out on a limb and said that inventories would turn upward within 2 weeks. While they did fall again last week, I am sticking to my prediction that they will turn back upward this week.

    I guess you have to be right eventually, but you’re beginning sound like a broken record 🙂

  17. We have the production numbers, the import numbers, and how much inventories fell. I think from that we can figure out how much production then came from U.S. refineries.

    Production last week was 8.777 and blending components imports 0.675. This would give (8.777 – 0.675) / 0.8825 = 9.180 MMB. Taking an average of the last 4 weeks imports gives a slightly higher figure of 9.250 MMB.

    So, maybe something around 9.2 MMB is maximum refinery gasoline production capacity? It would be useful to have such a number to get an idea of what utilization and imports would be required to meet demand at (say) last year’s levels.

    I take it exports also have to be considered as a ‘drain’ on domestic supply. In May 2005 and 2006 finished gasoline exports were roughly 170,000 bpd, not an insignificant amount in the present circumstances, but unfortunately I don’t think there’s any way to get hold of this year’s export number.

  18. “I guess you have to be right eventually, but you’re beginning sound like a broken record :)”

    I think you misunderstood what he said. He didn’t predict the same for next week. He said that he made the prediction two weeks ago that production would turn up within 2 weeks. Last week it didn’t, but he stuck to the prediction through the release of yesterday’s numbers. What you seem to be implying is that after this week he repeated the prediction. I don’t think that’s the case.

  19. I second that. One prediction doesn’t amount to a broken record. I don’t understand that comment. I think carbonsink just didn’t understand what was said – smiley face or not.

  20. Please understand, I mean no criticism of Robert. He is a thousand more times knowledgable about these issues than me, but I think his two week prediction is three weeks old now.

  21. I think his two week prediction is three weeks old now.

    I thought that was probably what you meant as well. Anyway, you would be wrong. Check. It was a prediction covering 2 weeks. It was not extended to a third, which was your obvious implication.

    Cheers, Robert

  22. My apologies. I interpreted Robert’s original prediction as meaning gasoline inventories would turn upwards in the next two TWiP releases. Well, two TWiPs have come and gone and inventories are still falling.

    So, if inventories fall again next week is your limb officially broken?

  23. I interpreted Robert’s original prediction as meaning gasoline inventories would turn upwards in the next two TWiP releases. Well, two TWiPs have come and gone and inventories are still falling.

    That interpretation is correct. Three weeks ago, following the inventory report, I said inventories would turn up within 2 weeks. After week one, they didn’t turn up, but I said that I maintain my prediction. Note that this was still the prediction I had made the previous week.

    The second week was this week, and inventories didn’t turn up. So, my limb did break. But what I didn’t do is then again say “inventories to turn up next week.” Because it would indeed be as you say: I have to be right sooner or later. So, pointless to make the prediction again.

    Cheers, Robert

  24. Screw the gasoline, I want to know about climbing out of a helicopter underwater 🙂

    Not as bad as I thought. I am going to write up something on it. My class of 16 did have 3 casualties, though. Only 13 of us finished once the helicopter exercise started. One showed up drunk; one walked off after being dunked, and one walked off before being dunked.

    Cheers, Robert

  25. I’ll be pedantic and rational on the subject of prices and peak oil as well.

    The thing to remember is that predictions of peak oil all rely on inductive logic. As such it is inappropriate to talk about “proofs” at all. There are merely strong and weak cases of induction.

    And at that level it is obvious that high prices are more supportive of a “peak oil” case than are low prices.

    While they do not “prove” they lend support, just as low prices would not “disprove” but would undermine support.

    – odograph

  26. Gasoline inventories up 400,000 barrels this week.

    Does this mean we are out of the woods for the U.S. driving season?

Comments are closed.