Corrosion in Underground Storage Tanks (USTs) Storing Ultra-Low Sulfur Diesel (ULSD): EPA Research on Risks, Prevalence, Causes, and Next Steps

Presented at the National Tanks Conference

September 16, 2015 Ryan Haerer

EPA Office of Underground Storage Tanks

Outline What is ULSD Corrosion? ◦ The problem ◦ Impacts ◦ Potential causes

What do we do with the results? ◦ Educate ◦ Address issues ◦ Update standards and practices ◦ Further research into prevention

and treatment

EPA’s Research Study ◦ Overview ◦ Key Findings ◦ Conclusions

Corrosion in USTs Storing ULSD Reports began around 2007 Internal metal components – often STP shaft Severe and rapid onset Yet unidentified cause Extent not fully known Appearance different and impacts more severe than corrosion in sump spaces of USTs storing gasoline/ethanol blends

What do we know about the possible cause?

Reports of corroded metal equipment in vapor space of USTs storing ULSD – first in 2007

Two changes to fuel supply around same time ◦ Reduced sulfur content in diesel beginning 2006 (500ppm LSD to 15ppm ULSD)

◦ Increased production and use of ethanol and biodiesel*

2012 Hypotheses Investigation by Clean Diesel Fuel Alliance of 6 UST storing ULSD * Energy Independence and Security Act of 2007 expanded Renewable Fuel Standard, setting volumetric blending targets for renewable fuels

Corrosion Risks to the Environment – Exposed Metals in the Vapor Space

Release prevention equipment could fail to function ◦ Corrosion on flapper valves could restrict movement and allow an

overfill ◦ Product level floats get stuck on corroded shafts and fail to signal a

rising product level, fuel release, or water infiltration

◦ Ball float valves – ball or cage may corrode

◦ Line leak detectors failing performance testing at higher rates

Some Observed Corrosion Examples –

Corrosion Risks to the Environment – Bottoms of Tanks

Metal components could corrode completely and possibly release fuel to environment

◦ ULSD prone to collect water and sludge in bottom of tanks

◦ Some jurisdictions seeing much higher rates of bottom failures of

primary walls of double-wall steel tank bottoms since ULSD

◦ Single-wall tanks possibly leaking undetected

Some Observed Corrosion Examples

Costs of Metal Corrosion for Owners

Increased pace of filter changes

More frequent servicing of equipment

Shorter lifespan before replacement of equipment

Findings from Clean Diesel Fuel Alliance 2012 Hypotheses Investigation

◦ Not conclusive, but suggested microbiologically influenced corrosion (MIC) a possible cause worth further research

◦ Microbes feed on ethanol present in ULSD, creating acetic acid

◦ Ethanol possibly entering ULSD through switch-loading of trucks wherein a gasoline-ethanol load is followed by a diesel one, or by diesel and gasoline-ethanol blend UST’s sharing the same vent line.

Ethanol can be converted into acetic and butyric acid (and possibly into glycolic acid)

Could other factors be in play? ◦ Glycerol can also be converted to corrosive acids

◦ Propionic acid presence in 2012 suggests this possibility

◦ Glycerol (possibly) in biodiesel, biodiesel possibly in ULSD* ◦ Allowable concentration from production, or

◦ Out of specification biodiesel

◦ Likely that a combination of factors involved

*ASTM D975 allows biodiesel to be blended into ULSD up to 5% without being labeled biodiesel

Glycerol into glyceric, lactic, and propionic acids

2014-2015 Study on the Corrosion of Metal Components in USTs Storing ULSD

Environmental Protection Agency Office of Underground Storage Tanks (OUST)

Battelle Memorial Institute

UST owner volunteers and industry partners, esp. CRC Diesel Corrosion Panel – could not have made the study happen without them! Thank you!

Provided critical site access for real-world data collection Provided data input on tank history and management practices Critical review of study design and draft reports

What did we set out to accomplish with the 2014-2015 Research Effort?

For continuity- ◦ Build on previous research and help figure out

how to address the problem ◦ Allow others to build on what we find

To better understand the extent of the problem and potential risks identified in the limited reports we’ve heard

Identify correlations among UST systems with severe or minimal corrosion

Definitively pinpoint a specific cause – every UST is unique

Identify specific solutions to the problem

Research designed: This research was NOT intended to:

Identify Volunteers for a Diverse UST Sample Population

• 10 geographic clusters • 42 sites • 24 fiberglass, 16 steel, 2 steel coated • 8 of 10 have steel and fiberglass in cluster

• 8 owners

• Government, retail, fleet • Single and multiple site • Large range of fuel throughputs and suppliers

• Diverse USTs

• 1 – 29 years in service • 5,000 to 20,000 gallons in capacity • Different product storage histories • Various approaches to maintenance

0

4

8

12

16

20

5,000 6,000 7,000 8,000 10,000 12,000 15,000 20,000

# of

UST

s

Tank Capacity (gallons)

42 USTs by Capacity and Material

Steel Fiberglass

Collect Data on UST Conditions at Each Site

Inspect with internal tank video

Collect samples: • Vapor • Fuel • Water bottom

Collect information on maintenance, throughput, fuel supply, biocide use, etc.

Sample Analyses

FUEL

Fuel Analysis Methods Method Identifier Determination of Water in Petroleum Products, Lubricating Oils, and Additives by Coulometric Karl Fischer Titration (Procedure B)

ASTM D63048 Water Content

Determination of Density, Relative Density, and API Gravity of Liquids by Digital Density Meter ASTM D405210 Density

Acid Number of Petroleum Products by Potentiometric Titration ASTM D66411 Total Acid Number

Determining Corrosive Properties of Cargoes in Petroleum Product Pipelines NACE TM-17212 Corrosion Rating

Particulate Contamination in Middle Distillate Fuels by Laboratory Filtration ASTM D621713 Particulates Determination of Biodiesel (FAME) Content in Diesel Fuel Oil Using Mid Infrared Spectroscopy (FITR-ATR-PLS Method)

ASTM D737114 Biodiesel Content

Flash Point by Pensky-Martens Closed Cup Tester ASTM D9315 Flashpoint

Determination of Free and Total Glycerin in Biodiesel Blends by Anion Exchange Chromatography

ASTM D759116 Free and Total Glycerin

GC-MS Full Scan Lab In-House Method Unknowns of Interest

Determination of Total Sulfur in Light Hydrocarbons, Spark Ignition Engine Fuel, Diesel Engine Fuel, and Engine Oil by Ultraviolet Fluorescence

ASTM D545317 Sulfur Content

Electrical Conductivity of Aviation and Distillate Fuels ASTM D262418 Conductivity

Determination of Short Chain Fatty Acids by Gas Chromatography-Mass Spectrometry (GC-MS) Lab In-House Method Acetate, Formate, Propionate, Lactate, Glycerate

WATER

Water Bottom Analysis Methods Method Identifier Determination of

Ion Chromatography (IC) for short chain fatty acids Modified EPA 300

Acetic, Formic, Propionic, Lactic Acids

IC Test for Free Glycerin Lab In-House Method Glycerin

Determination of Dissolved Alkali and Alkaline Earth Cations and Ammonium in Water and Wastewater by Ion Chromatography

ASTM D691919 Cations (Sodium, Calcium, Magnesium, Potassium, Ammonium) and Anions (Chloride, Sulfate, Nitrate and Fluoride)

pH (Electric) EPA 150.120 pH Conductance (Specific Conductance, umhos at 25°C) EPA 120.121 Conductivity Nonhalogenated Organics Using GC/FID SW846 8015B22 Ethanol and Methanol

FUEL

Vapor Analysis Methods Method Identifier Determination of

Ullage % Relative Humidity Hygrometer used per

manufacturer instructions

% relative humidity

Carboxylic Acids in Ambient Air Using GC-MS ALS Method 102 Acetic, Formic, Propionic, and Butyric Acids

Determination of Lactic Acid in Ambient Air

Modified NIOSH 7903 Lactic Acid

VAPOR

Assess and Categorize Corrosion Coverage

Based on protocol developed by CRC Diesel Performance Group – Corrosion Panel

3 assessments of coverage on STP shaft ◦ Minimal as < 5% ◦ Moderate from > 5% to < 50% ◦ Severe as > 50%

If in unanimous agreement, UST was categorized If not in agreement, discussed and considered overall condition of UST equipment and categorized by panel vote

Observed Corrosion Examples – Example Of An UST System with a Fiberglass Tank With Minimal Corrosion Coverage (Installed 2003 And Age Of Filter < 1 Month).

Example Of An UST System with a Fiberglass Tank With Severe Corrosion Coverage (Installed 1986)

Example Of An UST System with a Steel Tank With Minimal Corrosion Coverage Example (Installed 1994)

Example Of An UST System with a Steel Tank With Severe Corrosion Coverage (Installed 1992).

Observed Corrosion Examples –

More Corrosion Examples – STP Shafts

Drop Tubes

Tank Walls

Ball float cages

ATG floats/shafts

3 8 7 4

9 11

0

4

8

12

16

20

Minimal Moderate Severe

42 USTs by Corrosion Category and Material

Steel Fiberglass

Looking at the Data

Analyzed data according to corrosion category

Identify potential predictive characteristics

Identify any trends with respect to corrosion development

A note about findings: Please do not cite. This is preliminary data from the report which has not yet completed peer-review. This is presented only to share what we initially observed. Only the final peer-reviewed paper should be cited as an official source of information from the study.

Key Findings

Corrosion appears to be very common. 35 of 42 USTs judged as moderately or severely corroded.

Likely affecting same or higher proportion of total ULSD UST population

– emergency generators and owners of single UST systems were not represented in this sample.

Corrosion could pose a serious risk to integrity or functionality of metal UST system components.

No ‘smoking gun’ correlation among corroded tanks identified.

Gasoline found in all 42 samples, and ethanol in 38 of the samples.

Assumptions

Microbiologically influenced corrosion (MIC) likely responsible for the corrosion.

◦ Food sources (ethanol or glycerol) were present

◦ All water bottoms found acids that could result from microbial consumption of ethanol or glycerol

◦ Conditions good for microbial growth

Eliminating water is recognized as a key factor in preventing this corrosion.

From ASTM D975: X6 | MICROBIAL CONTAMINATION

X6.1 Uncontrolled microbial contamination in fuel systems can cause or contribute to a variety of problems, including increased corrosivity and decreased stability, filterability, and caloric value. Microbial processes in fuel systems can also cause or contribute to system damage.

X6.2 Because the microbes contributing to the problems listed in X6.1 are not necessarily present in the fuel itself, no microbial quality criterion for fuels is recommended. However, it is important that personnel responsible for fuel quality understand how uncontrolled microbial contamination can affect fuel quality.

X6.3 Guide D6469 provides personnel with limited microbiological background an understanding of the symptoms, occurrences, and consequences of microbial contamination. Guide D6469 also suggests means for detecting and controlling microbial contamination in fuels and fuel systems. Good housekeeping, especially keeping fuel dry, is critical.

Next steps Educate owners about the risks.

Use available treatments or preventions if your system is affected

Remove water, conduct testing, tank lining, oxygen displacement, biocides, corrosion inhibitors, others.

Explore changes to technical standards and operating procedures

Support continued research on causes and solutions

Thank You for Listening!

Ryan Haerer EPA Office of Underground Storage Tanks

Haerer.Ryan@epa.gov 703-347-0151

Thanks to:

Anne Marie Gregg, Battelle For her efforts on the study and materials for

this presentation.

GreggA@Battelle.org 614-424-7419

Please reach out to EPA to share input, ideas, information or with any questions.