HomeProducts & ServicesPublishingInTechLevel 1Communities

1 June 2005

Clean fuel regulations cross contaminate

The unlikely mix of diesel and jet fuel sparks industry action.

By Rodney Spitler and Alberto Pasquale

Low sulfur fuel specifications are presenting tremendous challenges to the refining industry.

Distribution of low sulfur motor fuels by pipelines presents yet another set of challenges.

Ultra low sulfur diesel (ULSD) will be limited to 15 parts-per-million sulfur (ppm S) by weight (wt/wt) by mid 2006. ULSD will travel through a common pipeline system with 1,000 to 3,000 ppm S (wt/wt) distillates, such as jet fuel.

The opportunity for sulfur contamination of the ULSD diesel is enormous. Rapid, precise, and online determination of total sulfur will provide a means for sulfur contamination detection in the fuel distribution system.

Technicians and engineers came up with a unique approach to perform laboratory simulations of pipeline interfaces between high and low sulfur fuels. This response time and analytical performance of an online method for determination of total sulfur in refined product pipelines follows.

Pipeline operators anticipate

The U.S. Environmental Protection Agency (EPA) and other environmental regulatory agencies have promulgated regulations requiring dramatic reductions in motor fuel sulfur content. The regulations will most immediately influence highway motor fuels.

In the U.S., sulfur content of highway diesel fuel will go from 500 ppm S (wt/wt) to 15 ppm S (wt/wt) by mid 2006. The average sulfur content of gasoline will be limited to 30 ppm S (wt/wt) beginning early 2006. Off-road, marine, and locomotive diesel fuel will contain 500 ppm S (wt/wt) to approximately 3000 ppm S (wt/wt). Jet fuel will also contain several-thousand-ppm sulfur for the near future.

The low sulfur diesel and gasoline are clean fuels. The vast majority of clean fuels and high sulfur fuels emanate from refineries to product terminals in refined product pipelines. A pipeline operator may send a batch of clean fuel directly behind a batch of high sulfur fuel. These fuels will mix in the pipeline to form a pipeline interface or transmix. Detection of an ultra low sulfur diesel, ULSD, and off-road diesel interface will likely require online sulfur measurements. The primary property difference between high sulfur and ULSD is sulfur content. Therefore, other methods of interface detection, such as, density measurement are unreliable.

The clean fuel regulations have substantially increased the sulfur content differential between highway fuels and various off-road fuels. Today, the sulfur content of highway diesel averages approximately 300 ppm S (w/w) while off-road diesel can be as high as 3000 ppm S (wt. /wt.)—a 10:1 difference. The sulfur content difference between ULSD, 15 ppm S (wt. /wt.), and off-road diesel could be as high as 200:1. The average sulfur content of low sulfur gasoline will be 30 ppm S (wt. /wt.); the sulfur content differential between it and off-road fuels could be as high as 100:1. The potential for sulfur contamination of clean fuels will increase dramatically as the clean fuel regulations take effect. In addition to interface detection, pipeline operators anticipate that online sulfur concentration measurements will be necessary to prevent sulfur contamination of clean fuels from pipeline dead legs and other unexpected contamination sources. The challenges of the online sulfur concentration measurements at refined product pipeline terminals are many. The primary challenge is to provide a very fast response to changes in sulfur concentration. Changes in sulfur concentration must login in 60 seconds or less.

Ultraviolet fluorescence

Reliable detection of pipeline interfaces or sulfur contamination will require an online sulfur analyzer with a rapid response time. Analyzer response time must directly relate to a known change in sulfur concentration. The analyzer's initial response to a sulfur concentration change, X, and the analyzer's final response time (time to stabilize at a new sulfur concentration), Y, must be considered.

The test apparatus built enabled the calculation of the actual sulfur concentration as a function of time to determine the online analyzer's response time. This test apparatus has allowed precise determination of the analyzer's response to simulated pipeline interfaces and "slugs" of pipeline sulfur contamination.

The output of the binary blending pump piped to the inlet of Thermo Electron's online total sulfur analyzer, the SOLA II. This analyzer measures motor fuel total sulfur content by sample combustion, then by pulsed ultraviolet fluorescence spectrometry (PUVF).

New samples load to the analyzer once every 15 seconds for refined product pipeline applications. Each 15-second sample mixes with air and burns. All motor fuel sulfur compounds convert to sulfur dioxide (SO₂) during the combustion process. The PUVF spectrometer measures the amount of resulting SO₂.

The result of this process is a batch-total-sulfur-concentration output representative of pipeline conditions. Essentially, the analyzer is reporting the rate of total sulfur concentration change in the pipeline. This unique method for the online analysis of total sulfur enables the rapid detection of pipeline interfaces and "slugs" of pipeline sulfur contamination.

The binary blending pump can simulate a variety of potential sulfur concentration profiles that might show up in a refined product pipeline. The actual sulfur concentration as a function of time can be determined from the blending pump's program. The following response times were determined:

a) Response when sulfur concentration is changing from high to low.

b) Response when sulfur concentration is changing from low to high.

c) Response to a "slug of sulfur contamination."

The pump's dead time—the time required for a new sample to travel from pump to the analyzer's injection valve—was determined as follows:

R_PS = R_S + R_P 

Where:

R_PS = Total observed response time

R_S = Analyzer response time (X or Y)

R_P = Time required to deliver new sample from pump to injection valve

Assuming R_P is a linear function of the total pump flow then for pump flow rates of 2.5 ml/min and 5.0 ml/min we have:

at 5.0 ml/min:

R_PS1 = R_S + R_P 

and at 2.5 ml/min:

R_PS2 = R_S + 2R_P 

By substitution and rearrangement, we have:

R_P = R_PS2 – R_PS1 

The blending pump delivered a 5.0-minute linear change from 322.1 ppm S to 6.5 ppm S at total flow rates of 5.0 ml/min and 2.5 ml/min. At 5.0 ml/min, R_PS1 was 80 seconds. At 2.5 ml/min, R_PS2 was 100 seconds. The estimated pump "dead time," R_P, is 20 seconds.

Executive summary action

To summarize, the response time of our creation, an online total-sulfur-analysis method, evaluated using a binary blending pump, which enabled the simulation of pipeline interfaces and sulfur contamination, took place and was successful.

Proven were initial response times of from 20-60 seconds. The variability of response times is likely dependent upon the rate of pipeline-sulfur-concentration change and the time at which the sulfur concentration change goes to the analyzer.

The analyzer introduces a new sample for analysis once every 15 seconds; therefore, if a pipeline sulfur-concentration change occurs immediately following the introduction of a new sample, one can expect a minimum analyzer response delay of 15 seconds.

We conclude that the analyzer's response time is faster when the rate of pipeline-sulfur concentration change is greater. These results indicate this unique batch method of total sulfur analysis can provide an excellent representation of refined product pipeline-sulfur concentration profiles.

Behind the byline

Rodney Spitler is petroleum market specialist at Thermo Electron Corporation. He is an ISA member. Alberto Pasquale (alberto.pasquale@thermo.com) also works at Thermo Electron and has over 20 years experience in the design and programming of embedded systems. He presently concentrates on the design of firmware for process analyzers.