Refining 101: Winter Gasoline

One of the things I really like about The Oil Drum is that I learn something new here almost every day. So, I thought I would return the favor by explaining in a bit more detail something you may not know.

Every year in late summer, you will start hearing references in the media about the conversion to winter gasoline, such as the following (originally in the Bradenton Herald, but the link is long dead):

Motorists can thank a mild hurricane season in the Atlantic for the lower gas prices, according to the American Automobile Association.

Other factors include the end of the summer driving season and a cheaper winter fuel mix.

Gas stations sell a special, more expensive fuel blend during the summer to cut down on smog during hot months. Stations nationwide will start selling a less-expensive winter fuel blend Friday, which could lead to even lower prices, analysts said.

So what does this mean, and why does it make winter gasoline less expensive?

A Primer on Gasoline Blending

Gasoline is composed of many different hydrocarbons. Crude oil enters a refinery, and is processed through various units before being blended into gasoline. A refinery may have a fluid catalytic cracker (FCC), an alkylate unit, and a reformer, each of which produces gasoline blending components. Alkylate gasoline, for example, is valuable because it has a very high octane, and can be used to produce high-octane (and higher value) blends. Light straight run gasoline is the least processed stream. It is cheap to produce, but it has a low octane. The person specifying the gasoline blends has to mix all of the components together to meet the product specifications.

There are two very important (although not the only) specifications that need to be met for each gasoline blend. The gasoline needs to have the proper octane, and it needs to have the proper Reid vapor pressure, or RVP. While the octane of a particular grade is constant throughout the year, the RVP spec changes as cooler weather sets in.

The RVP is the vapor pressure of the gasoline blend when the temperature is 100 degrees F. Normal atmospheric pressure varies, but is usually around 14.7 lbs per square inch (psi). Atmospheric pressure is caused by the weight of the air over our heads. If a liquid has a vapor pressure of greater than local atmospheric pressure, that liquid boils. For example, when you heat a pot of water, the vapor pressure increases until it reaches atmospheric pressure. At that point, the water begins to boil.

In the summer, when temperatures can exceed 100 degrees F in many locations, it is important that the RVP of gasoline is well below 14.7. Otherwise, it can pressure up your gas tanks and gas cans, and it can boil in open containers. Gas that is boiled off ends up in the atmosphere, and contributes to air pollution. Therefore, the EPA has declared that summer gasoline blends may not exceed 7.8 psi in some locations, and 9.0 psi in others.

A typical summer gasoline blend might consist of 40% FCC gas, 25% straight run gas, 15% alkylate, 18% reformate, and 2% butane. The RVP of the gasoline blend depends on how much of each component is in the blend, and what the RVP is of each component. Butane is a relatively inexpensive ingredient in gasoline, but it has the highest vapor pressure at around 52 psi.

In a gasoline blend, each component contributes a fraction to the overall RVP. In the case of butane, if there is 10% butane in the blend, it will contribute around 5.2 psi (10% of 52 psi) to the overall blend. (In reality, it is slightly more complicated than this, because some components interact with each other which can affect the expected RVP). This means that in the summer, the butane fraction must be very low in the gasoline, or the overall RVP of the blend will be too high. That is the primary difference between winter and summer gasoline blends.

Why Prices Fall in the Fall

Winter gasoline blends are phased in as the weather gets cooler. September 15th is the date of the first increase in RVP, and in some areas the allowed RVP eventually increases to 15 psi. This has two implications for gasoline prices every fall. First, as noted, butane is a cheaper blending component than most of the other ingredients. That makes fall and winter gasoline cheaper to produce. But butane is also abundant, so that means that gasoline supplies effectively increase as the RVP requirement increases. Not only that, but this all takes place after summer driving season, when demand typically falls off.

These factors normally combine each year to reduce gasoline prices in the fall (even in non-election years). The RVP is stepped back down to summer levels starting in the spring, and this usually causes prices to increase. But lest you think of buying cheap winter gasoline and storing it until spring or summer, remember that it will pressure up as the weather heats up, and the contained butane will start to vaporize out of the mix.

And that's why gasoline prices generally fall back in the fall, and spring forward in the spring.

So just curious, but how much difference is there in the blends in various regions of country?

For instance living in Houston we have fairly hot summers, and mild winters (I think the first snow fall I've seen since living down here for 18 years was just 2 years ago on Christmas Eve).

Given Houston's relatively stable temperature range, does Houston see as much of this cheaper blend than say New York?

Also while it may not be a good idea to save Winter Gas for summer, are there disadvantages to using Summer Gas in winter?

Like Robert, I once was a refinery scheduler and blender. My job was to write and run computer models (linear programming) to figure out the best crudes to process and to come up with the recipes for various gasoline and fuel blends.  

Two problems arise from using winter and summer blends out of season. If you live in a very cold climate and tried to use summer gasoline, the lower RVP of the gasoline will make it more difficult to start the engine. Liquid fuels must be vaporized before they will burn. This isn't as much of a problem with fuel injection as the fuel is atomized and sprayed into the piston.  On the intake stroke, the pressure in the cylinder is reduced to below the boiling point of the fuel, so some of the fuel quickly turns into a vapor. On the compression stroke there is not enough time for the vapor to recondense. Cold start problems were common on carbuerator equiped engines.  

Using winter gasoline in the summer could cause something called vapor lock, where the fuel vaporizes in the fuel lines before reaching the carb or fuel injectors.  Again, this used to be a pretty common problem when cars had mechanically driven fuel pumps that sucked the gasoline from the tank before forcing the fuel into the carbuerator. Cars today are equipped with high pressure electric fuel pumps mounted inside the gas tank.  These pumps "push" the liquid fuel towards the engine.

I would assume that is why those of us who have tractors have to put stablizers in our tanks for the winter. What exactly does these additives do?

thanks for the lesson.

Gasoline & diesel stabilizers are added for a different reason. To understand the need for them you would need a lesson in basic refining.

Crude oil is mixture of hydrocarbon molecules that have from 4 carbon atoms to 50 or more. Most of the molecules in crude are long single chains of carbon atoms with hydrogen atoms attached.  Called "alkanes" or "parafins" these molecules have no double chemical bonds.

A very simple refinery would process the crude by distillation alone.  The distilling process would seperate the crude into smaller fractions of mixtures of similar hydrocarbons.  These fracions are called "straight runs" by the blender. Because they come straight from the crude without further processing. The problem is that distillation alone doesn't produce very much of the gasoline and diesel components.  A typical barrel of crude might only produce 1/2 barrel of these valuable products. The rest of the crude would be in very light components (C5 pentanes and lighter) and low value heavy fuels (C50 & greater).  

To create more gasoline and diesel components, the refiner uses further processing with heat, pressure, and exotic catalysts to break down longer chain hydrocarbons into shorter ones with the desired properties. These processes might include cat-cracking, vis-breaking, hydrocracking, coking, etc.  

During this process, some of the hydrocarbons form double bonds (a carbon atom making 2 bonds with an adjacent carbon atom).  These compounds are called "unsaturates" because they are not fully saturated by hydrogen atoms.  Double bonds are more reactive and less "stable" than single bonds. In the presence of oxygen, like in a gas or diesel can that is opened frequently, the double bonds react to form longer chain hydrocarbons, cyclic compounds and other undesired forms that can lead to gums and heavy deposits.

Adding stabilizers to the diesel or gasoline the stabilizer reacts with the double bonds to keep the heavy compounds from forming.  

You can see national (by state) requirements for RVP for conventional gasoline here (PDF, 7 pages), and a map of variations by county (included reformulated gasoline) here.

The latter map gives you a sense of the multiple recipes for gasoline blending that refiners complain about, and you can see why California is virtually an island market and has to depend on its own refineries for gasoline production.

Thank you, Robert, for sharing the information.  This is the beauty of TOD, we all learn something every day because those with specialized knowledge help us to understand, and the collective fund of information keeps growing.  
Great article.   Love to learn.

As I understand it,  they do something similar for diesel - in winter there is an anti-gel agent?   Is that right?

If you were to store diesel,  would it be better to store winter diesel?

On the topic of storage,  I have read that you can store diesel for a long time (20 yrs) if conditions are right,  is that correct?

It's all about population!


From this page, there are two grades of diesel for vehicles, No. 1 or No. 2.  No. 1 is lighter, has a lower cloud point, and has less energy than No. 2.   Winter diesel can be a blend with more No. 1 and less No. 2 than summer diesel.  Winter diesel can also have additives to reduce wax issues.

Thank you RR.  I've always wondered about this.
same here!
Ah, so no political conspiracy?


My wife is convinced that the upcoming election is behind the drop in prices. Now, in addition to the ending of the SDS, mild hurricane season, repair of Katrina damage, and somewhat lessening of geopolitical tensions I can tell her about RVP values too.

It's interesting that a side effect of changing blends might be to increase the probability that incumbents, regardless of party, are re-elected in November elections.

If I had the time and inclination, tracking the timing which these changes occurredwould be an interesting way to test how influential gas prices are to incumbent reelection in a way that gets around the problem that politicians may be manipulating gas prices.

Anybody think there is an RVP blend effect?

I never knew the RPV vapor lock connection.

Very informative, Robert.  Keep up the great work.

An interesting little aside...with fewer storage tanks and bulk storage along the pipelines and implementation of a JIT system, the amount of time to empty the storage tanks of of the summer blend for the winter blend (and visa versa) is much shorter.  

Right now, stocks of winter blend are being sent out the pipelines to various terminals for distribution after September 30th.  And terminals are depleting their stocks of summertime blends.  

OPEC Lowers Fourth-Quarter World Oil Demand Forecast

By Stephen Voss

Sept. 15 (Bloomberg) -- OPEC, which pumps 40 percent of the world's oil, lowered its forecast for fourth-quarter and 2006 demand, citing a slowdown in the world's largest economies.

The Organization of Petroleum Exporting Countries cut its outlook for fourth-quarter world demand to 85.6 million barrels a day, down 300,000 barrels a day from its previous forecast a month ago. OPEC also reduced demand forecasts for the current quarter, and for 2007, according to an e-mailed monthly report today.

``The slowing economy along with high oil prices has impacted oil demand growth, leading to a downward revision'' of North American 2006 demand of 100,000 barrels a day, the report from the OPEC secretariat in Vienna said. ``The summer peak for gasoline was not as strong as expected.''

Oil intensity, or the amount of crude used to produce a unit of economic output, will probably continue to decline as a result of high prices, the group's report said. OPEC this week decided to maintain its current oil production target, as analysts expected. The organization hasn't changed the quota for more than a year as oil prices rose to record levels above $78 a barrel this summer.

``In the long run, strong prices may intensify the downward trend in intensities by accelerating oil-saving and substitution effects,'' the group said in the report. Substitutions include power generators using natural gas rather than fuel oil.

The producer group cut estimates for 2006 and 2007 world demand by 100,000 barrels a day. It still expects world consumption to increase by 1.5 percent or 1.3 million barrels a day next year, which is similar to the current production of Algeria.

OPEC Production

OPEC's 11 member nations pumped 29.789 million barrels of crude a day in August, up 90,400 barrels a day from July mainly because of an increase in Venezuela, the report said, citing its survey of ``secondary sources,'' which include estimates by news agencies and analysts.

Excluding Iraq, which was no quota, the other 10 members pumped 27.73 million barrels a day, which was close to the collective target of 28 million barrels.

Supply from non-OPEC countries was forecast at 55.4 million barrels a day for 2006 and 57.5 million barrels a day for 2007, unchanged from OPEC's previous estimate.

``One of the issues we have is that non-OPEC supply is going to increase very much next year,'' Iranian Oil Minister Kazem Vaziri-Hamaneh told reporters in Vienna on Sept. 12.

The changes in the world demand forecast led to corresponding revisions to OPEC's estimate of the demand for its crude oil, which was lowered by 300,000 barrels a day for the fourth quarter, to 28.9 million barrels a day, OPEC's monthly report showed. The so-called ``call on OPEC'' crude was lowered by 200,000 barrels a day for both 2006 and 2007.

Reflecting seasonal changes in world demand and a gradual rise in supply from outside the exporter group, the calculated demand for OPEC crude is expected to peak at 29.2 million barrels a day in the first quarter of 2007, before dropping to 27 million barrels a day in the second quarter.

To contact the reporter on this story: Stephen Voss in London at

Last Updated: September 15, 2006 08:19 EDT

I wonder when OPEC will begin citing "the development of oil alternatives" as being the reason that "no one is buying" their crude and that the rate of these "alternatives" will happen to match the expected production decline due to depletion.
And that's the true tale of why gasoline prices fall back in the fall, and spring forward in the spring.

This refers to gasoline relative to oil (crack spread) but not directly to absolute gasoline prices as they rely on crude oil as the primary input.

If gasoline prices truly were that predictable to go down every fall, that would already be in the market by July or August - there are many years in the last decade when prices increased in the fall so this phenomenon is a tendency not a rule. I think oil has gone down 6 of the last 10 years from Sep to Dec - I'll have to check on unleaded.

Regarding fall vs spring pricing, there exists an active futures market for oil and natural gas out for 60 months. For unleaded gasoline however, the futures only go out until December of this year.

This refers to gasoline relative to oil (crack spread) but not directly to absolute gasoline prices as they rely on crude oil as the primary input.

If gasoline prices truly were that predictable to go down every fall, that would already be in the market by July or August - there are many years in the last decade when prices increased in the fall so this phenomenon is a tendency not a rule.

Hey Nate, I made it to California. Just checking in from the business center in the hotel.

Not all of the inputs come from crude oil. That is part of the point. In the winter, the fraction of butane goes up, meaning that gasoline supplies increase every fall. Whether price rises for some other reason, supply always increases as the vapor pressure goes up because of the ability to put butane in.

That's all for now. Got to run.


This web site has a serious problem. It has absolutely wicked dead spots that seem fairly easy to fix. It's Friday about 3pm. Granted, nothing much is going on in the world of oil. But the site is clocking over 500 hits an hour. This is near peak. But nobody is posting. I tried throwing a Bloomberg article out there to get some talk going. Nothing. I'm not usually here this time, but when I am, it's pretty obvious. Are you telling me only the non-English writing Kuwaitis are logging on. It's noon in California. What's the problem? Don't tell me it's the Pacific Ocean. When people start logging on when they know their friends are gonna be on. That's bad. That creates a polarized community.
Alot of people log on to read, not post. If they dont have anything intelligent to say, it suffices to read the knowledge of others. If everyone who logged on made posts let alone repeated postings, it would reduce the EROI for everyone. And youre Bloomberg post might have been more appropriate in the open thread, rather than one on blended winter gasoline.
Except that this doesn't make any sense. Everyone decides to stop writing and read at the same time? Not. It's because they get burnt out. It has to do with timing and spacing of threads. I have no idea who's listening, but I've expressed this sentiment before. Yours is exactly the second piece of feedback I've gotten. Thank you. We're a bit short, but you are slated to get 93 black-eyed virgins, depending on rate of decline factors. Trust me on this one.
No, you're not the first to notice this, CEO.  It's especially the case on Fridays though, which may just be people being fried from the week, etc.
When I referred to Nate as the second, I was implying Jack as the first. And this was in relationship to me whining.

As far as "me" noticing "this" first - I have no idea. But I'm glad "you" as in "the" Professor Goose is at least responding to my concerns. Thank You. I couldn't ask for anything more. I mean that. I'll shut the hell up now.

I have about four years of experience using scoop based discussion sites. Once threads go beyond ~150 posts they become hard to read. There is also an ego based disincentive as posts that are buried deep in a discussion thread are less likely to get attention. There is a similar disincentive for threads that are over ~6 hours old as they are less likely to get new/repeat readers.

I don't have ideas for fixing this beyond enabling the diary/story feature for individual users (ala dailykos) which I don't think the admins are interested in. Allowing personal diaries would space things out a bit and could serve as a partial replacement / enhancement for the daily open threads. On the other hand personals diaries become competition for the overall narrative of the site which I imagine the admins aren't interested in.

Personally I prefer this site as it is - this topic is extremely prone to conspiracy theory and emotional hysteria, thus to be a bit elitist I don't want the "rabble" in a position to drive the narrative more than they already do.

Hey, rabble here.


Well the "irony" is that I'm part of the rabble too. I guess I hate freedom.
I've wondered if two threads might be appropriate at times.  One aimed a little more towards technical aspects, and another for a little more speculative discourse.  But I can't even begin to predict how something like that would turn out.
I can assure you that I am not "the" anything...  !laugh!  

Don't shut up!


The price of crude has dropped a lot recently.  Are there any projections on how long this drop will last?  Or is this like trying to predict the weather?  It's funny how many fields tend to be horribly inaccurate in the short term but incredibly accurate in the long term.

"`Shoulder Season'

``We're still in shoulder season,'' Sieminski said. ``Gasoline season is over and heating demand has yet to pick up.''  (...)

U.S. gasoline demand normally declines after the Labor Day holiday in early September. Global fuel use peaks during the Northern Hemisphere winter. Heating oil consumption jumps as temperatures plunge and furnaces are stoked in Europe, North America and eastern Asia."

Are there any projections on how long this drop will last?

Im reasonably certain the market is waiting for me to liquidate my last long oil and gasoline futures positions and then it will head higher. I'll keep you posted.

Gas prices below $2 in some places
Interesting article...and interesting post Kingofkaty!  

Amazing how a little knowledge can topple the house of cards built buy the Big oil conspiracy nuts! (but don't expect the major media to help enlighten the public on the real reasons for energy price fluctuations)