On U.S. Finished Gasoline Stocks and Gasoline Refining Capacity

This is a guest post by the thegoodshipkship.

Several TOD commenters expressed concern recently regarding U.S. finished gasoline inventory decreases as evidenced by the weekly EIA reports. Total gasoline stocks are presently at the bottom range of the last five years as seen in FIGURE 1 (under the fold) and apparently crude prices increased partly as a result of this information. However, one can also see that total stocks were above five year highs over several periods and, as seen in FIGURE 2 (also under the fold), crude stocks are above the top range of the last five years by some 20 mb. At first glance there seems little to be concerned about...


FIGURE 1: 2006 U.S. Total Gasoline Stocks including Blending Components. Graph from the EIA Weekly Gasoline Report.

FIGURE 2: 2006 U.S. Crude Oil Stocks. Graph from the EIA Weekly Crude Oil Report.

However, if one subtracts out the blending components, we see finished gasoline stocks are well below the 5 year range as seen in FIGURE 3.


FIGURE 3: 2006 U.S. Finished Gasoline Stocks (Conventional + Reformulated). Source: EIA Historic Weekly Data.

Reformulated gasoline imports dropped from around 200 kbd in January 2006 to 0 kbd by this past summer and it would seem this is the main cause of the 2006 decline. From the EIA report (pdf) on the switch to ethanol from MTBE, one might expect that the finished gasoline stock decline in 2006 is a one-off as the industry adjusts. However, FIGURE 4 shows this is not a trend isolated to 2006.


FIGURE 4: Average U.S. Finished Gasoline Stocks 2001 - 2006. Source: EIA Historic Weekly Data.

Not only has demand exceeded supply in 2006, but for the last six years as well. Since 2001, stocks have declined at approximately a rate of 7.3 mby or 20 kbd. Note that this is a rather small rate (we will return to this shortly). The EIA data show that this trend started at least in 1990. Since then, stocks have declined slowly and recently at a faster rate.


FIGURE 5: U.S. Finished Gasoline Stocks 1990 - 2006. Source: EIA Historic Weekly Data.

Before jumping to conclusions, let us also examine the other side of the story -- U.S. finished gasoline demand and utilization of refinery capacity. FIGURE 6 shows the finished gasoline balance between demand, imports, exports, and production for 2006 (Data for exports was available only through September).


FIGURE 6: Demand/Supply and Export/Import Balance. Sources: EIA Historic Weekly Data and EIA Monthly Export Data.
The weekly export rate (in kbd) was approximated as equal to the monthly rate (in kbd).

The balance should be roughly equal to draws and inputs to finished gasoline inventories. A back of the envelope calculation using the average of the blue series predicts a draw from stocks of approximately 20 mb as of September '06 from January '06. FIGURE 3 shows this is in rough agreement to reported finished gasoline stocks as expected. The pink series whose average is +65 kbd shows that if we were not exporting finished gasoline (to the tune of ~139 kbd) stocks would be increasing. Apparently we're exporting our finished gasoline stocks.

A brief glance at refinery capacity utilization is also revealing. The EIA data show on average in 2006 utilization was at 89.24% with an operable capacity of approximately 17,300 kbd which translates into approximately 9,861 kbd of operable capacity for finished gasoline (over 1mbd in excess capacity). This assumes ~57% of the capacity can go to finished gasoline. The EIA says as much as 60% is possible. In order to cover the 74 kbd average draw from stock an increase of utilization to 90% on average would have been required (+0.76%). Similarly, the average utilization from 2001 to 2006 would have to increase from 91% to 91.2%, an increase of one fifth of one percent, in order for it to have covered the 20kbd on average draw from stocks mentioned earlier. All things being equal, had utilization averaged 91.4% over the 2001 to 2006 time frame there would be a record surplus of gasoline stocks of approximately 44 mb above 2001 (~90 mb above current levels).

FIGURE 7 shows 2001 to 2006 average U.S. finished gasoline production capacity ( assumed 57% of operable capacity), U.S. demand, and the three cases corresponding to 91, 91.2, and 91.4 % utilization for U.S. production plus imports minus exports. The difference between the demand curve and the yellow curve should be equal to daily stock draws and puts. The difference between the yellow (actual, 91%), aqua (91.2%), and maroon (91.4%) series is the difference between what would be in 2006 record lows and record highs in finished gasoline inventory. The pink series shows the ample production capacity.


FIGURE 7: 2001 - 2006 U.S. Demand and Operable Capacity.
Sources: EIA Historic Weekly Data, EIA Monthly Export Data,
and EIA Operable Capacity and Utilization.
The weekly export rate (in kbd) was approximated as equal to the monthly rate (in kbd).

Conclusions

While finished gasoline stocks have declined to below 5 year lows the fact that stocks have been declining in a similar fashion since 2001 says that 2006 is not particularly different from the immediate past. Were the U.S. headed towards a shortage reducing the export of finished gasoline would remedy it. However, the extremely slow rate of 20 kbd at which stocks have been declining since 2001 and the marginal adjustment of refinery utilization (assuming the production capacity is not a fiction) that would arrest that decline suggest it is not one caused by either refinery bottlenecks or supply shortages. The obvious conclusion is that stocks have been allowed to diminish.

Folks, consider this a reminder to positively rate these articles (using the icons under the tags in the story title) at reddit, digg, and del.icio.us if you are so inclined. Also, don't forget to submit them to your favorite link farms, such as metafilter, stumbleupon, slashdot, fark, boingboing, furl, or any of the others. These posts are a lot of work, and the authors appreciate your helping them get more readers for their work however you can.
The EIA data show on average in 2006 utilization was at 89.24% with an operable capacity of approximately 17,300 kbd which translates into approximately 9,861 kbd of operable capacity for finished gasoline (over 1mbd in excess capacity). This assumes ~57% of the capacity can go to finished gasoline. The EIA says as much as 60% is possible.

First of all, very nice work. Good graphs, and good data interpretation.  

The problem with going to 60% going to gasoline is that diesel demand is incredibly high, and diesel prices are very high. So, everyone is trying to make as much diesel as they can, which takes away from gasoline production. Shifting more production to gasoline would cause a diesel emergency.


RR,

As I have discussed in my prior Diesel posts, the projections have been that Diesel fuel would be higher right now to amortize the cost of the ULSD (Ultra Low Sulfur Diesel) switch, and the cost of the hydrogen generators to achieve this. What seemed to be a surprise to the Diesel consumer is that ULSD requires from 25% to 40% more hydrogen to achieve the sulfur standard than it did under the old 500ppm standard.  This is drawn from natural gas, so if natural gas prices go up, Diesel prices go up, whether crude oil went up in price or not.  Also, if lower cost and more readily available "sour" crude is used, the consumption of hydrogen aka natural gas will have to rise.

For those who said we could never build a hydrogen economy, guess again, because we are building one whether we like it or not, by way of the hydrogen generation at refineries to cleanse Diesel fuel to make it legal.  Even the DOE and EPA didn't hazard a guess in public about how much valuable natural gas would be consumed for this, but it is given as the BIGGEST single ongoing expense in the ULSD process, so it must be considerable.

I have stated on several occasions what a debacle I think this will become for the Diesel industry, as more and more customers turn to LPG and CNG engines.  My view is that we have seen Diesel engine peak, and will now move on to clean quiet fuels, since we would have to use clean quiet fuels to make dirty loud power to stay with Diesel, so why do it?

Roger Conner  known to you as ThatsItImout

 

Roger,

I don't think we have seen the diesel peak. Despite the great progress nat. gas engines have made over the years, they still cannot match diesels for fuel efficiency unless they are run pilot injection with a diesel cycle (in other words diesel engines running on natural gas)It is in fact my opinion that using valuable nat. gas neat in engines is counterproductive and it is better used for heat, cogeneration , fertilizer, fuel cells etc. where much higher efficiency of use is possible.

Using some gas to lower sulphur and allow continuation of use of the most efficient transportation prime mover readily available (diesels) makes more sense than using 100% neat nat. gas at worse overall efficiency using Otto type cycles. Look at the transit bus industry where in some cases the nat. gas buses are up to 50% less fuel efficient than the diesels.


Sheldon,

You may be right, it's not a point I would fight over.  Price will determine everything.  I am thinking as a Diesel customer, and paying a premium of a quarter a gallon for fuel adds up FAST (and we are paying that much here in central KY, I don't know what the rest of the country looks like).  The other shoe is still to fall, when we go from 80% having to comply to all 100% in 2010, and when heavier sour crude starts coming through the refineries.  I have a contact in the refinery trade who says this is already a struggle for the oil companies, and they are battling dislocations throughout the petrochemical industry due to the cost of natural gas.

 Further, let's not forget the Nox standards which this year even the Diesel miesters themselves at Volkswagen could not make, and Benz has been coming up with some odd and fancy technical footwork to try to stay in the Diesel market.  You hit the nail on the head, though, nat gas has low heat density.  It makes one really look at LPG, but I do not know if there is enough of that around for a massive conversion to it.  

Interesting times, indeed!  
Roger Conner  known to you as ThatsItImout

That's all??  

Here in Los Angeles, regular unleaded is running about $2.30-2.40/gal while the CHEAPEST diesel is now about $3.15 and I have seen $3.49 at some stations.  

I had been wondering what was causing this huge difference. I had assumed it was the winter heating oil demand, or a refinery shutdown in CA, but apparently there's more than that going on.


Indeed.

Most of us oil burner's were caught off guard by this one! :-(

As mentioned, CA has to face a very stiff Nox emission law to boot, much worse than the rest of the country.

RC  known to you as ThatsItImout

you need to go look at the wholesale prices for gasoline vs diesel before you draw conclusions.

First, the wholesale price of gasoline vs LS diesel is only about 40 cts if you believe the DOE.  The rest of that difference is retailers wanting a big premium to keep pumps dedicated to low volume sales.  Same thing on premium.  the pump differential can be 20 cts when the wholesale regrade is a nickel.

http://tonto.eia.doe.gov/dnav/pet/pet_pri_spt_s1_d.htm

Secondly, West coast refiners have a well insulated mkt.  there are few import facilities and few independent wholesaler/retailers.  So even if a foreign cargo headed for NYH is diverted it can be very hard to find a place to put it or a buyer to take it in.  Can be very painful speaking from experience.

This is where allowing the refiners to consolidate in th e 90's really screwed over the consumer.

RR/Others,
Is the extra tonnage of NG required really significant (e.g. 1% of US NG consumption, 10% ??)

The product must be H2S? What is done with that gas? Is it flared or can it be burnt to recover energy?

Thanks.

The H2S is burned in a reducing environment to form some SO2 and then passed over a series of catalyst beds (Claus Unit) that convert and recover as molten elemental Sulfur.  

Refineries cannot flare this gas because of SO2 limits

Most of the heat of combusting the H2S is recovered as low pressure steam.

The problem is getting rid of the sulfur, as the crude get more sour and the more sour nat gas we are pulling out of the ground, the sulfur market cannot take up all the excess sulfur.

Thanks HiD and Slatz. Sounds like NG pricing for ULSD or any LSD is not material. They probably use that low pressure steam for some process purpose where they may have burnt Natural Gas so perhaps not much of an NG loss which would be burnt for the same purpose anyway.
the amount of nat gas going into hydrogen production is trivial compared to heating demand.  Keep in mind much of the the h2 used in refineries comes from reforming gasoline to bump up octane.  also if nat gas is expensive, low quality naphtha is easily reformed into hydrogen as a replacement..  You just have to swap out the catalyst

Moreover, just meeting 500 ppm for road diesel already took most of the sulfur out of diesel pool.  that spec was 2000-3000 ppm even as late as the 80's.  

Good work, Profgoose.

I can foresee a situation for gasoline similar to what we now have for natural gas: In natural gas, our production capacity is basically running at a constant rate, and the inventory is allowed to build up during low demand seasons, and drawn down during peak demand.  

You can envision a situation where instead of investing in additional capacity, the producers invest in additional storage space. Then, they run the refineries at 92% year round, and satisfy the peak demand out of storage and also out of imports. This would allow even better capital efficiency than we have now, but would also mean that the effects of downtime and/or reactor monkey business would be a lot more serious.

Obviously, we are not at that paradigm yet.

As of last fall, the European producers were running in the mid-80% range for capacity utilization, an had on the order of 50 days usage in storage, so this is still a readily available source of supply if necessary.

Also note that import availability will be a lot price driven, and also exchange-rate driven. At the moment, neither pricing nor exchange rate are favoring increased imports, as reflected in the reduction of imports over the last few weeks.


I'm a bit confused naively we equate gasoline stocks with crud e supplies but because we import so much gasoline its not a simple relationship.

I think the key issue I don't understand is when do you import and when do you make gasoline from crude ?

Maybe Robert can shed some light on these decisions.

And generally we import gasoline from Europe right ?

If so are their gasoline imports on the west coast and if so from where ?

If I had to guess I think the current situation is caused by the recent fall of the dollar vs the euro. We slowed imports and gasoline stocks declined but have not made it up yet by internal refining or can we not make it up ?

I guess at the end I don't understand why we are in our current situation of low gasoline stocks low imports and high crude stocks.

I can answer part of it. Yes, gasoline inventories are typically from Europe, although I think Brazilian ethanol gets classified as gasoline. Not sure if they get imports on the West Coast.

The reason we are in this situation is that imports dried up as the gasoline price fell. Refinery utilization is the bottleneck, so more crude doesn't help - we can't refine it fast enough. If prices get a bit higher and stay there you will see gasoline imports pick back up.

Gotta run. Cowboys are about to play. That's all from me tonight.


Thanks ...

But imports dried up because gasoline feel and gasoline fell because ??

And why buy all that crude of you can't refine it. Maybe because you believe prices are going up on crude ?
Or probably more correct this is leftovers from the non hurricane season ?

In anycase it seems to be a unatural situation that won't last forever. And I see two reasons for gas prices to go up. We run low on crude and import more or we import more gasoline.

A bit bothersome is I assume its at least three weeks if not more to ship gasoline from Europe so unless those tankers are on the way now things may get tight.

Yes, there are gasoline imports to the west coast. They averaged just 981 kbbl per month from April through September, less than a day's regional production. The linked table shows the most significant sources during that period seemed to be South Korea and Singapore.
think this is wrong:

Yes, gasoline inventories are typically from Europe,

Bet we get more from Venz.

and I'd be wrong.  I'm out of date (retired from the game in mid 90's):

http://tonto.eia.doe.gov/oog/info/twip/twiparch/060706/twipprint.html

"So, everyone is trying to make as much diesel as they can, which takes away from gasoline production. Shifting more production to gasoline would cause a diesel emergency."

Fuel Twitching on the producer end of the market as well as on the consumer end.  But going forward there is less wiggle room left for twitching.  Chaos in the markets maybe.

Thanks Prof. Goose.

If you take a look at petroleum and product exports from the US, you will notice that they have increased significantly over the last year or so:

http://www.bea.gov/bea/newsrelarchive/2006/trad0906.pdf

Check out page 14.  Petroleum and product exports have, in fact, almost doubled over the last (please note however that is in terms of $s and not barrels).

With the fall in Cantarell's production, and Mexico's admission that it has increased gasoline imports, I suspect this has much to do with them.  

Gasoline futures crashed from $2.30 to $1.40 in the past few months and have recently recovered to $1.60 - $1.70.
Perhaps the large crude stockpile has been putting pressure on prices.

A longer term chart shows an upward, though volatile trend.

Note: unlike crude oil and natural gas futures, gasoline only trades for this month and next month delivery, as opposed to any month in the next 5 years.

RBOB (unleaded gas) contracts have open interest thru Dec 2007.
The unleaded gasoline contract (HU) trades Jan and Feb only. RBOB (reformulated gasoline) goes out a year.
My understanding is that the HU contract will cease to exist after Feb. It seemed like you were implying that gasoline normally only trades two months in advance.
until the RBOB contract, it did only trade two months in advance. So - the Feb contract and no more Hu, huh?
that's inaccurate.  HU Contracts always went out 5-6 monthss at a minimum.  I thought it was 12 bck in the 80's but could be wrong.  There wasn't any liquidity in the way outers, but they were there.

HU is indeed dead.  that's why GS pulled HU out of the GSCI an starting in Sept.  And they didn't just replace it all with RB as they were likely concerned re the number of contracts of gasoline they were holding.  Looked to be >20% of the open interest which is ridiculous for a formulaic index with openly known roll dates.

I see up to 12 contracts per day as far back as the mid-90s in my data. This continued until the recent switchover period.

As a side note, when the commodity index was recently changed and unleaded gas was dropped, the rest of the group (lt. crude, brent, and heating) picked up the difference -- and those that got higher weightings see higher inventories currently. In other words, the inventories appear to go with the available hedging contracts. Imagine that!

I don't know what the difference is between gasoline capacity and finished gasoline supplied, but this link takes you to the latter:  http://tonto.eia.doe.gov/dnav/pet/hist/wgfupus24.htm

It shows that November, 2006 finished gasoline supplied is at about 9.3 mbpd (with finished gasoline stocks of 110 million barrels at the end of November)  versus 8.7 mbpd for November, 2001 (with finished stocks of 162 million barrels at the end of November, 2001).  

November consumption since 2001 is up by 5.7%, while stocks have fallen from 162 to 110, so relative to consumption, finished gasoline stocks have fallen from 18.6 days to 11.8 days, a drop of about 37%.  

I assume that there may be a statistical anomaly of some sort here as a result of the switch to ethane,  but we see a similar, but not so pronounced drop in total gasoline stocks, about a 12% drop in total gasoline stocks, relative to daily consumption.

In any case, IMO, we are going to have to sharply bid up the price of Total Petroleum imports, if we wish to continue consuming petroleum products at our current rate.  

Should read:  I assume that there may be a statistical anomaly of some sort here as a result of the switch to ethanol.  
Could all this be due to "market economy": drive down stocks to achieve higher gasoline prices?
First of all, I appreciate the editors deciding to post this.  Second, I plan to add comments and reply to comments but it will have to wait until later.  Too busy!
Regards,
thegoodshipkship
Hi Thegoodshipkship,

Thank you very much for your post. I think your analysis gives some very good clues about the current situation. You leave me wondering about some questions though :

If the current reduction in finished gasoline stock is voluntary, what can be the motives for doing so ?

You say that all that one of the solutions for the situation would be to cut exports. But aren't these exports, which go mainly to mexico and a bit to canada, part of a general exchange scheme and such not so easily removed ?


I know we import a lot of gasoline how does that fit in is it counted as stocks ?

Also considering the size of tankers would not a significant amount of gasoline be in flight so to speak in tankers headed for US ports ?

Finally although we treat the US as a single market Its my understanding that the west coast and east cost are not linked leading to at least two markets. Could the gasoline draw down simply be supply problems on the west coast from issues in Alaska ?

imports are counted as stocks once the bbls are onshore and the b/l finalized.  they are counted in the API/DOE stats at that point.  

there really isn't all that much in transit.  Typical imports are say 1 MMBD and sailing time from the biggest exporters such as the Venz (5 days to NYH) or NWEurope (10 days ish from memory) doesn't add up to all that much stock in transit.

West Coast (PAD V) is very separate from the rest.  Almost all imports head to PAD I (NY area).

This gas draw is just economics.  It's much more profitable to make heat at the moment and let gas stocks slide a little.  There's little likelyhood of a gasoline demand spike in winter so why make surplus?  Once Jan comes and heating oil production season begins to wane, you'll see refiners switch back to max mogas to tank gasoline for the spring shutdown season and summer demand season.  The contango is already pretty huge which will motivate folks to stock up on gasoline.

Goodshipkship, thank you for great work!
Couple of comments and questions, though. Refineries are always shutting down to do maintainence,switchovers and upgrades. Does the 91-92% utilization you quoted make any allowance for these shut ins? It strikes me that we are prety much at peak apacity.
  The military has huge gasoline and diesel useage, with much of it overseas. Is this coming from the domestic stocks? Does the hydrocarbons used in Iraq count towards domestic fuel capacity and useage?
Does the 91-92% utilization you quoted make any allowance for these shut ins? It strikes me that we are prety much at peak apacity.

A typical refinery has uptime in the 90-93% range every year due to maintenance, unplanned shutdowns, etc. If you take all refineries as a whole, utilization is probably never better than 92 or 93%. In fact, the average for 2004 was 91.6%, and the average to date in 2006 is 89.2% (more spring maintenance than normal this year as refineries delayed fall 2005 maintenance because of Katrina). For all practical purposes, we are running at close to 100% of what we can run.

However, refineries are always debottlenecking, so capacity will increase over time. The big question is whether it will keep up with demand.

Thanks, Robert, that's about what I thought about refinery utilisation. What worries me is that any kind of disruption-a hurricane, revolution in Mexico, Venezuella, Saudi Arabia or Nigeria ect. leaves us extremely vulnerable with ony 11-12 days of supply of gasoline and another 22-23 days of crude. In the mean time Yergin and Lynch and XOM offer us the vaium of saying everything's rosy for the next 30 years.
  Of course with you I'm preaching to the choir! Halleluja brother, they'll be gas in the sky when you die! My fathers house has many mansions, and no gas lines! Blessed are the poor in transportation fuel, for their spirit will have wings when they die!
  Quotations are from a Gnostic Gospel, the Book of Daniel Yergin, Peter Jackson translation
 

 Couple of questions for the experts:

  1) What is the BOE of LPG from a transfuel producing standpoint?

  2) If China is only using 40% of their oil for transfuel, then how are their refineries set up?


  Apologies. Should read: NGL
Prof. Goose,
Do you know how much natural gas goes into making a gallon of gasoline? I was surprised that oil refineries are the biggest user of natural gas in California. Bigger even than power than power plants, which I find hard to believe.
Hi

If you look at the numbers on the TWIP page, you'll see that the drop is actually just an accounting error.

Finished RFG stocks are down 21m barrels on a year ago - but blendstocks are 22m barrels up on a year ago.

Most of the difference is in PADDs I and III - the two coasts, where the cities are (and where most RFG use is). They're blending ethanol, you may recall.

Because of the phase separation problem, RBOB and ethanol are only blended together at the last possible moment, and are generally bunkered separately. Which means you can't report them as finished gasoline.

So the graph that the EIA publishes is a bit misleading...

ITs VERY misleading. They shoudl highlight on their website that they used to show apples and oranges together but now show them separately so if you are looking at fewer apples, there is a reason.
http://www.peakoil.com/article21474.html