Automatic Parking Assist standard on every Chevrolet SS Sedan
Daimler to invest $2B in Sindelfingen Plant by 2020

NIST study suggests severe corrosion in underground gasoline storage tanks may require component replacement sooner than expected; 500K USTs in US

Optical micrographs of severe corrosion on steel alloy samples exposed to ethanol and acetic acid vapors—conditions typical of underground gasoline storage tanks—after 355 hours, 643 hours, and 932 hours. Source: NIST. Click to enlarge.

In recent years, field inspectors in nine states have reported many rapidly corroding underground gasoline storage tank (UST) components such as sump pumps. These incidents are generally associated with use of gasoline-ethanol blends and the presence of bacteria, Acetobacter aceti, which convert ethanol to acetic acid, a component of vinegar. Corrosion can result in failures, leaks and contamination of groundwater, a source of drinking water.

Following up on the inspectors’ findings, a National Institute of Standards and Technology (NIST) laboratory study has demonstrated severe corrosion—rapidly eating through 1 millimeter of wall thickness per year—on steel alloy samples exposed to ethanol and acetic acid vapors. Based on this finding, NIST researchers suggest gasoline stations may need to replace submersible pump casings, typically made of steel or cast iron, sooner than expected.

Such retrofits could cost an estimated $1,500 to $2,500 each; there are more than 500,000 underground gas storage tanks around the country. The study is in press in the journal Corrosion Science.

A NIST study found that corrosion may pose a hazard at underground gasoline storage tanks at filling stations. The study focused on sump pump components, especially the pump casings (labelled #3 in graphic), which are typically made of steel or cast iron. Graphic: EPA. Click to enlarge.

The NIST study focused only on sump pump components, located directly below access covers at filling stations, just above and connected to underground gas storage tanks. The sump pumps move fuel from underground tanks to the fuel dispensers that pump gas into cars. These underground tanks and pipes also may be made of steel and could be vulnerable, too.

We know there are corrosion issues associated with the inside of some tanks. We’re not sure, at this point, if that type of corrosion is caused by the bacteria.

—NIST co-author Jeffrey Sowards

Much of the US fuel infrastructure was designed for unblended gasoline. Ethanol is now widely used as a gasoline additive due to its oxygen content and octane rating. A previous NIST study found that ethanol-loving bacteria accelerated pipeline cracking. (Earlier post.)

For the latest study, NIST researchers developed new test methods and equipment to study copper and steel alloy samples either immersed in ethanol-water solutions inoculated with bacteria, or exposed to the vapors above the medium—conditions mimicking those around sump pumps. Corrosion rates were measured over about 30 days.

The NIST study confirmed damage similar to that seen on sump pumps by field inspectors. The worst damage, with flaky iron oxide products covering corrosion, was found on steel exposed to the vapors.

Copper in both the liquid and vapor environments also sustained damage, but corrosion rates were slower. Steel corroded very slowly while immersed in the liquid mixture; the NIST paper suggests bacteria may have created a biofilm that was protective in this case.

Although copper corroded slowly—it would take about 15 years for 1.2-millimeter-thick copper tube walls to develop holes—localized corrosion was observed on cold-worked copper, the material used in sump pump tubing, NIST co-author Elisabeth Mansfield noted. Therefore, stress-corrosion cracking is a concern for bent copper tubing because it would greatly reduce tube lifetime and result in leaks.

The NIST test equipment developed for the study could be used in future investigations of special coatings and biocides or other ways to prevent sump pump failures and leaks.

NIST held a workshop in July 2013 on biocorrosion associated with alternative fuels.


  • J.W. Sowards and E. Mansfield (2014) “Corrosion of copper and steel alloys in a simulated underground storage tank sump environment containing acid producing bacteria,” Corrosion Science In press, corrected proof available online. doi: 10.1016/j.corsci.2014.07.009



Another of the 1001 reasons why we should switch to BEVs and FCEVs?

Concurrently, we should also switch from CPPs and NGPPs to clean RE sources.

Both are possible at much lower (per capita) NET cost than many posters would suggest.


I always said since the beginning that ethanol was a very bad idea for all kind of reasons, first it is made with foods, it has lower btu content and now we know that it is corrosive.


Looks like there are problems with ethanol all down the line;


Make bio synthetic gasoline using thermochemical processes, not ethanol.

Roger Pham

Easy fix: keep gasoline ethanol-free, and ethanol in separate container, and blend them together at the pump.
This will give great deal of flexibility to FFV that can run on higher ethanol content. Future cars can be designed with a separate ethanol tank with a separate pump for blending at the injector. No ethanol needed at low loads or at high engine speeds. Then, more ethanol will be mixed in with increasing load to prevent knocks. Engine compression can be raised to improve efficiency.

For cars with a separate ethanol tank, the ethanol can be hydrous ethanol which cost less energy to make and result in higher efficiency, while engine can run cooler at high loads with more power with injection of water and ethanol mixed with gasoline. WWII aircraft got power boost via water and alcohol injection permitting higher boost pressure.

Roger Pham

Instead of ethanol from grain, methanol, aka wood alcohol, can be made much more efficiently and cheaply from cellulosic wastes via thermal process with catalysts. Methanol is more corrosive than ethanol, however, in cars that has a separate methanol-conpatible tank, the corrosiveness to the engine can be avoided by running the engine without any methanol prior to shut down. In fact, that should be how future ICEV should run. Save the grains (corn) for people and livestock.


I was catching up on the GCC right-column articles, and only a few moments ago read the RFS response on E85 availability ("Renewable Fuels Association says DOE is under-representing E85 stations"). Now, according to AFDC and commenters around all the greentech boards/blogs, there is no problem whatsoever with higher ethanol blend --- the issue is market interference by Big Oil.

As a minority investor from '08-'12 in a chain of convenience store/refilling stations in Northern Illinois, I was part of the decision to avoid offering E85. Virtually nobody on this board will believe it, but we received no threats at all from the major wholesalers from whom we purchased fuels, nor from XOM (who sold us the "core" locations from which the chain was built).

Nope, it was feedback from one of our other board members whose experience in equipment rental showed significant damage from the use of E85. We also had no way to effectively obtain indemnification for misuse of high ethanol fuels. Add in the the cost of UL-rated components, insurance, etc.

NACS members have been preaching about this hazard for years. Plenty of people use E85 in cars that are rated for it, but most just don't want it. Over and over I hear that negative research about E85 is Big Oil backed... believe what you want. I guess they also poisoned the storage tanks of retailers (right after they finished destroying all the Tesla Free Energy machines and buried the last of GM EV1s)



The Open Fuel Standard would require FFVs to take ethanol AND/OR methanol. It has been in Congress for five years with no result. I will leave to people to connect the dots to see why.

California ran hundreds of cars on M85 for years with total success. The auto makers recently said they would comply with OFS at no additional cost to the buyer. In other words a "slam dunk".

Now what methanol would do to fuel station pumps is another matter. Assuming you could retrofit a tank to handle methanol you could have blender pumps to dispense M5 to M85, depending on the vehicle. There are more than 100,000 fueling stations in the U.S., static inertia is a force against change.

Roger Pham

Since methanol is corrosive to the engine, it would be best to blend the methanol at the engine fuel injector instead of at the gas station pump. In this way, the proportion of methanol can be adjusted depending on engine load and speed. The engine should run on mostly gasoline prior to shut down to purge methanol out of the engine.

Though there are 100,000 gas stations, only less a 1000 would be needed to have a station within 5-7 mile driving distance in urban and suburban areas. A typical gas station has 8-24 gasoline pumps, but initially only one methanol pump would be needed, so the retrofit cost can be minimal.


IMO people don't want two tanks and two fuels they have to put in the car. Car makers have said M85 is NO problem for the tank, pump, lines, injectors nor engine.

The blended M85 in the tank is PROVEN to work in trials using lots of vehicles over many of years. The one drawback is shorter range. If you have a 400 mile range per tank you might get 300 miles with M85, but the Fuel is under $3 per gallon.

Jim McLaughlin

Biocides have been often needed in 100% petroleum diesel for a long time if I recall, you can buy the stuff at WalMart. Biocides have been around in hydraulic fluid for many decades, even metal working fluids have the stuff. Why is this a surprise that biocides might be needed in gas storage tanks too?

Even with a biocide added, E85 is probably still far less toxic to you the customer than 100% gasoline. Are pipes more important than you are?

Roger Pham

@SJC and Jim,
There are many reasons high-alcohol-content fuels not popular in motor vehicles, the following quote from Wikipedia:
"Both methanol and ethanol burn at lower temperatures than gasoline, and both are less volatile, making engine starting in cold weather more difficult. Using methanol as a fuel in spark-ignition engines can offer an increased thermal efficiency and increased power output (as compared to gasoline) due to its high octane rating (114[10]) and high heat of vaporization. However, its low energy content of 19.7 MJ/kg and stoichiometric air-to-fuel ratio of 6.42:1 mean that fuel consumption (on volume or mass bases) will be higher than hydrocarbon fuels. The extra water produced also makes the charge rather wet (similar to hydrogen/oxygen combustion engines) and with the formation of acidic products during combustion, the wearing of valves, valve seats and cylinder might be higher than with hydrocarbon burning. Certain additives may be added to the fuel in order to neutralize these acids.

Methanol, just like ethanol, contains soluble and insoluble contaminants.[11] These soluble contaminants, halide ions such as chloride ions, have a large effect on the corrosivity of alcohol fuels. Halide ions increase corrosion in two ways; they chemically attack passivating oxide films on several metals causing pitting corrosion, and they increase the conductivity of the fuel. Increased electrical conductivity promotes electric, galvanic, and ordinary corrosion in the fuel system. Soluble contaminants, such as aluminum hydroxide, itself a product of corrosion by halide ions, clog the fuel system over time.

Methanol is hygroscopic, meaning it will absorb water vapor directly from the atmosphere.[12] Because absorbed water dilutes the fuel value of the methanol (although it suppresses engine knock), and may cause phase separation of methanol-gasoline blends, containers of methanol fuels must be kept tightly sealed."

However, when used in vehicles in separate tank in variable proportion to boost octane rating to enhance both power and efficiency, and to reduce NOx and smog, methanol is great, so long as the engine is run on mostly gasoline before being shut off.

Don"t worry, the gain in fuel efficiency and power at lower fuel cost will motivate motorists to fill up the methanol tank after filling up the gasoline tank. Those who are easy on the gas pedal will not have to fill up the methanol tank that often.

Roger Pham

So, if the gas station would have a second container for alcohol, that container will have a thick plastic lining to separate the alcohol from the metal casing. The alcohol tank will be 1/10 the size of the pure gasoline tank, making retrofit much cheaper than having to replace the huge E10 tank. Even if the alcohol tank should leak, alcohols are far less toxic and polluting than gasoline in E10, and there only 1/10 the volume of E10.

Biocides add cost and can pollute the environment, and more chemicals to deal with.


What I get from this article is that some sump pumps may need to be replaced more frequently, at a trivial cost (around $2,000). Think of how much the fuel in the tank is worth and you will see that the pump pales in comparison.

The article is needlessly misleading when it mentions half a million tanks, because it's obvious that not all will be affected: some are diesel tanks, some are in areas that do not use ethanol.

Roger Pham

Well, not just the pumps casing that are corroded but the tanks and the piping also, although access to the pump is easiest so that was illustrated. Petroleum spill is very serious for the environment.


When you consider all the toxic sites with leaking gasoline tanks, it is a wonder. Methanol exists in nature, there are organisms in the soil that take care of that.


'Certain additives may be added to the fuel to reduce acidity'

In fact, that is what reagent buffering is all about. Add a weak acid such as gasoline to a stronger one such as ethanol, and the solution will generally be less acidic than either.

I can assure you that even gasoline will rot the metal of a tanker. I knew a welder who tried to repair the hole in a tanker and he gave up -- the material was as porous as a sponge. There is the problem of oxidation, which hygroscopically absorbed water can cause, but also of hydrogen embrittlement. Paraffinates can help keep the hydrogen out of circulation,and so can shorter aliphatics. But blending ethanol gasolines to avoid these problems seems to be in its infancy. Could hydrofurans be the answer?

The comments to this entry are closed.