Quick and Dirty Feedstock Characterization

Practical advice for cash-strapped community-scale biodiesel plants
By Christina Borgese and Marc Privitera | July 18, 2011

Feedstock characterization, process conversion and fuel finishing are the building blocks of biodiesel production. Of the three, feedstock characterization is usually underestimated. The lack of depth in the understanding of feedstock’s impact on the business plan has lead to challenges that might have been avoided for smaller scale producers. This article on feedstock is the first of three installments to discuss the building blocks of biodiesel production.

Nontraditional methods to process higher free fatty acid (FFA) feedstocks are more technically complex compared to traditional systems. High FFA feedstocks include yellow grease, brown grease, tallows, and algal oils. Yellow grease is primarily comprised of restaurant and cooking wastes. Brown grease typically comes from grease trap waste, dissolved air flotation  skimmings, agricultural spoils and meat cut waste. Algal oil is only now emerging as another feedstock. Tallows and rendered fats typically have a high existing market value. 

The National Renderers Association defines yellow grease as no more than 15 percent FFA and no more than 2 percent MIU (moisture, insolubles and unsaponifiables). The historical reference on FOG, Bailey’s Industrial Oil and Fat Products, defines brown grease as having an FFA level between 15 and 50 percent. There is much debate throughout the industry on how exactly to define brown grease. In reality the nomenclature is inconsequential. What really matters is the actual FFA and MIU content received at the plant and whether or not the system is capable of processing the material. A rookie mistake is negotiating a contract to buy yellow or brown grease without the needed characterization. 

The excitement of the green movement has enabled community-scale efforts to repurpose locally collected waste material. Concurrently, municipalities have been working to solve the problem of FOG continuing to create sewer system backups. Over the last 10 years local municipalities have enacted new regulatory practices for all FOG producers requiring increased collection and reporting. This has created a need for a community-based outlet for the collected material. Most commonly this material is land-applied or dewatered and disposed in landfills. Many small biodiesel producers have begun to look at this locally collected waste material as a feedstock. The characterization methods of that feedstock are critical to their success.

In every system from community-scale to major industrial operations, it’s important to characterize your inbound raw materials. Regardless of your processing capabilities, the critical feedstock measurements are moisture, insolubles, solid particle size, melt point curve, FFA and sulfur. The test methods to determine these characteristics need to be quick and reliable so that the material can be analyzed before it enters the feedstock storage tank. 

The first step in feedstock characterization is to identify a potential supplier and set up the collection of samples for a defined period of time to gather a statistically significant set of data for the proposed feedstock supply. This data set is usually comprised of a minimum of seven samples. To be statistically rigorous, the number of required samples is calculated using the precision of the measurement and the desired risk tolerance in the results. 

A common brown grease source is a restaurant grease trap. Material in grease traps is usually only 2 percent recoverable oil depending on the pumping frequency of the trap. This 2 percent, however, if allowed to enter the sewer system, is what causes all the problems. Time-honored practices in processing collected grease trap waste (GTW) focus on dewatering the material with a polymer then solidifying the remainder for land application or land fill disposal. The oil phase almost always becomes part of the solidified sludge and previously was considered a low-value material. In recent years new process technologies have been developed that can convert high FFA oil into biodiesel.

These new processing capabilities have placed a renewed focus on separating the oil from the solids in an effort to recover as much value from the collected material as possible creating new composting and energy producing opportunities. When polymer is used to bind the oil with the remaining solids, additional attention in sampling and analyzing the sludge is required for accurate characterization of the feedstock.  

There are current requirements for sampling grease traps. The challenge is getting a good representative sample. Sludge sticks (clear PVC pipes with valves on the bottom) are a usual method for grabbing a trap sample. If the trap is pumped on a frequent basis, it could turn out that the oil layer is indistinguishable and the ability to calculate the amount available for conversion becomes difficult. A more convenient way to gather a representative sample is to sample from the discharge of the vacuum trucks after it enters the first stage of the dewatering process, usually the pit, and after it is well mixed as a load.

The focus should be first on the total amount of oil in the collected material. A centrifuge method is normally used. The ASTM D2709 Standard Test Method for Water and Sediment in Middle Distillate Fuels by Centrifuge specifies to use a 100 milliliter (mL) sample spun at a centrifugal force of 800 for 10 minutes. This test is part of the B100 fuel specification. Adopted from this test is a centrifuge method that uses a 15 mL sample and provides a visual representation of the feedstock’s characteristic fractions. This grossly determines the volume percents of floating and heavy solids, free water and oil. The main drawback to this method is that it lacks the capability to test bound water in the oil phase. 

Now with your separated centrifuge sample, pull an aliquot of the oil layer and titrate for FFA. This titration method should be derived from ASTM D974, Test Method for Acid and Base Number by Color-Indicator Titration. Quickly summarizing, dilute your sample in isopropyl alcohol and titrate it against standardized KOH using phenolphthalein as an indicator. 

Even though you have pulled an aliquot from your oil layer for your FFA titration, most likely your oil still is wet to some extent. A very important detail in the measurement of the moisture in the oil is the difference between the free and bound moisture. Each play a role in the reaction inhibition and the formation of the undesirable byproducts, but many times the collection of the sample, the sample preparation, or the measurement method does not give an accurate representation of the moisture present. 

A Karl Fischer titration is the classic method for measuring total moisture. The equipment for a Karl Fischer moisture titrator typically costs $5,000 to $8,000. It is a sensitive machine that requires frequent calibration and maintenance and in the biodiesel world it is best used to measure moisture in fuel against a known standard. The variability of MIU in feedstock often throws a Karl Fischer off its measurement causing many man hours to recalibrate and clean. 

A less expensive but reliable method used to test the moisture in the oil is water bomb. This method uses a pressure-versus-time correlation from reacting the test method reagent with the sample in a calorimetric bomb reactor. This quick method can provide valuable insight into the bound moisture that the centrifuge method is unable to provide.

Another simple but less accurate method for measuring bound moisture is the hot pan test. In a lab hood while wearing proper protective equipment, heat an electric frying pan and gather about 10 mL of oil for a sample. When the pan surface is at a stable 220 degrees Fahrenheit, measuring with a noncontact temperature device, pour 1 mL of feedstock into the pan. If it bubbles like crazy, stop.

This indicates that you have significant bound water and need to process the feedstock further. If there are no visible bubbles, pour the remainder of the material onto the surface of the pan. If the increased volume in the pan also shows no bubbles, then the feedstock is fairly dry. At this point remember to turn off the pan, or this test will turn into a smoke point test, which could be useful in other situations dealing with oil.  The hot pan test is especially useful as a visual aid to illustrate how much water is in the feedstock to those who may not understand the difference between free and bound moisture.

A more refined and detailed characterization should be completed offline in the lab. A temperature-versus-viscosity curve should be run. This will give an indication of how easily the water will separate from oil in the feedstock. As the oil heats up, the water will separate from the remaining liquid at different rates at different temperatures. This would not have been obvious in the centrifuge test previously conducted. The temperature-viscosity profile is critical to the proper operation of the disc stack centrifuge, which is usually the first of the plant operations in the separation of the oil from the water. If the viscosity it too high, the water is not able to move through the mass to coalesce with other droplets to become large enough to release from the oil. High viscosity oil will not be able to travel up the disc surface cavities to separate from the water, and the

water will be unable to release from the disc film and flow to the higher density collection port. 
The processing challenges of using wet, high FFA, brown grease feedstock can be overcome using proven processing methods including the traditional blending, acid esterification and even supercritical methanol technology. One detail of brown grease feedstock characterization that cannot be overlooked is sulfur content. Measuring sulfur in the feedstock is an expensive aspect that might not be in the planning budgets for many community-based and local-scale biodiesel producers. Brown grease is going to have a significant amount of sulfur. The technology and methods to remove the sulfur from the feedstock and ultimately the fuel is a difficult path to maneuver, as many of the required unit operations do not scale down economically from the normal large-scale refinery desulfurization systems that is a part of any dinosaur feedstock diesel operation.

Planning for sufficient analytical support for the needed sulfur disposition study throughout any nontraditional process system is critical as the 15 ppm specification is a difficult level to meet. Characterization methods are critical to benchmark the feedstock sulfur content.  Tracking the effectiveness of sulfur removal along with other process conversion technology details will be covered in future installments.  

Authors: Christina Borgese, Marc Privitera
Founding Engineers, PreProcess Inc.
(949) 201-6041

Array ( [REDIRECT_REDIRECT_STATUS] => 200 [REDIRECT_STATUS] => 200 [HTTP_HOST] => biodieselmagazine.com [HTTP_ACCEPT_ENCODING] => x-gzip, gzip, deflate [HTTP_USER_AGENT] => CCBot/2.0 (http://commoncrawl.org/faq/) [HTTP_ACCEPT] => text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 [HTTP_IF_MODIFIED_SINCE] => Wed, 17 Jan 2018 22:00:02 GMT [PATH] => /sbin:/usr/sbin:/bin:/usr/bin [SERVER_SIGNATURE] =>
Apache/2.2.15 (CentOS) Server at biodieselmagazine.com Port 80
[SERVER_SOFTWARE] => Apache/2.2.15 (CentOS) [SERVER_NAME] => biodieselmagazine.com [SERVER_ADDR] => [SERVER_PORT] => 80 [REMOTE_ADDR] => [DOCUMENT_ROOT] => /datadrive/websites/biodieselmagazine.com [SERVER_ADMIN] => webmaster@dummy-host.example.com [SCRIPT_FILENAME] => /datadrive/websites/biodieselmagazine.com/app/webroot/index.php [REMOTE_PORT] => 35084 [REDIRECT_QUERY_STRING] => url=articles/7928/quick-and-dirty-feedstock-characterization [REDIRECT_URL] => /app/webroot/articles/7928/quick-and-dirty-feedstock-characterization [GATEWAY_INTERFACE] => CGI/1.1 [SERVER_PROTOCOL] => HTTP/1.0 [REQUEST_METHOD] => GET [QUERY_STRING] => url=articles/7928/quick-and-dirty-feedstock-characterization [REQUEST_URI] => /articles/7928/quick-and-dirty-feedstock-characterization [SCRIPT_NAME] => /app/webroot/index.php [PHP_SELF] => /app/webroot/index.php [REQUEST_TIME_FLOAT] => 1521380800.815 [REQUEST_TIME] => 1521380800 )