bacOkay.  I have been asked many times what verifications are there that a particular fermentation culture is safe, “clean” or “pure”?  Whether its yogurt, kefir, sourdough, kombucha, buttermilk, cheese, or koji; the question remains. For the most part, natural fermentations are quite rigorous and the true culture organisms predominate.  If foodborne illness organisms get into the culture, it is likely they will not survive due to microbial competition from the culture.  However, studies of pathogens in fermentation cultures are few and far between.

About a decade ago I sent emails to about a dozen small business culture sellers and asked them what they did regarding the safety of their cultures they sell.   The answers I got back were quite impolite to anthropological, but none had any idea of the safety of their culture, its purity, or cleanliness. From the perspective of the FDA and the Food Code, culture manufacturers MUST be “inspected” or the culture would not be usable as an ingredient.  So, culture safety starts with: “Is the culture supplier an inspected facility?”  If not, stop, move on.  Ebay or other home seller marketplaces are not typically a place to find food safety.

The next question is “how does the manufacturer know the culture is “clean” or better “safe”?  As a Master Brewer, I can tell you that the moment a brewing yeast culture is contaminated, the beer will not be exactly as planned.  It could end up a total loss.  Larger culture companies understand the economics at hand and go to great lengths to ensure their cultures are clean, pure, and safe.  Most will provide a certificate of analysis including microbiological analyses.  For example a pure brewing yeast culture should have no bacteria, no coliforms, and no Staphylococcus aureus when tested.  Beer culture manufacturers will also ensure that there are no wild yeasts.  The presence of a wild yeast could make a great beer taste like a band aid.  Even the smallest culture manufacturer can send off some of their culture to a testing lab for basic microbial analysis.  But do they?

In the absence of lab data, the retail-foodservice operator could verify their culture fermentation by performing a test batch.  This is highly recommended simply because some microorganisms could die off in culture.  Acetobacter aceti is notorious for this.  For fermentations that produce high levels of acid, the acid can select for the intended culture and help eliminate spoilage and any possible pathogens.   At the same time as the acid level gets higher, more of the culture dies off.  What about back-slopping?  Back-slopping is the practice of using culture from a successful fermentation for the next fermentation.  If the current fermentation was successful in producing acids and flavors, then it is likely that the culture is “mostly” pure and a good candidate for back-slopping.  Back-slopping should be safe for a retail-foodservice operator fermenting foods using a HACCP plan, provided they provide an SOP describing where the original culture came from and how they determine the culture is maintained so that it is likely free of pathogens.

Bottom line – buy cultures from FDA or regulatory inspected manufacturers.  Ask for a certificate of microbial analysis.  Create a starter batch of product from fresh purchased cultures to see that the culture produces the intended product characteristics.  Use the “art” of fermentation to notice the nuances and slight differences that could mean the culture is not “clean and pure”.  Lastly, use excellent sanitation principles to minimize cross contamination or environmental contamination of the fermentation process. Re-use these starter cultures for as long as they demonstrate successful fermentation characteristics indicating they are “clean” and “pure”.

Large food fermentation culture companies: DSM, Danisco (including Kefir cultures), and Chr. Hansen.

Dry aging beef (and other meats) is said to enhance flavor and tenderness.  It is often used by upscale purveyors to distinguish their products from others.  Dry aging meat is essentially aging meat without protection is a refrigerator.  Cuts are stored separate from each other to permit air flow to dry them in 1-6 weeks.  Natural enzymatic reactions result in a change of flavor and an increase in tenderness.  Another major change is the loss of moisture resulting in a more concentrated flavor, but with a corresponding yield loss.  Several vendors are selling “dry age” bags.  The intention is to seal the surface of meat while permitting moisture loss (dry aging).  In experiments with the bags the moisture loss is lowered (higher yield) while still permitting the “dry aging” process leading to flavor and tenderness attributes.  The bags will also prevent yeast and mold buildup on the surface of dry aging meats.

cc 2008 Filipe Fortes
cc 2008 Filipe Fortes

Dry aging charcuterie (salami, pepperoni, procuitto) is also possible in dry age bags.  The main benefits would be allowing moisture loss and minimizing yeast or mold surface growth.  However, caution must be used since the traditional process for these meats may not be the same when bagged versus left open.

Regulatory concerns: The main concern regulators have for using these bags under the US FDA model food code is whether the bag results in a reduced oxygen packaging (ROP) process.  The bag is moisture permeable and may or may not be oxygen permeable.  Although the bags are quite thin, an oxygen transfer rate CoA (Certificate of analysis) is needed from the bag manufacturer demonstrating an OTR over 10,000 cc/m3/24h to exclude this process from HACCP (under the food code).  If the OTR is below 10,000 or unknown then a simple HACCP plan is required.  In this case storing the meat in the bag at ≤ 41°F for ≤ 30 days total (packaging to service) or ≤ 34°F for ≤ 60 days total (packaging to service) would be acceptable.  The 30 day limit is considered two barrier packaging (refrigeration and competitive bacteria) and a “safe harbor” process in 3-502.12 of the Food Code.  The 60 day shelf life would require a regulatory variance and is justified by limiting the maximum refrigeration temperature to 34°F using a 24/7 temperature datalogger.

Resources: check out a beef research dry aging white paper here.  No endorsements intended but here are some vendor websites:;;


Brian A. Nummer, Ph.D.
November 2004 (originally written for the NCHFP/University of Georgia)
Updated with Retail-foodservice options April 2015

SOP=standard operating procedure; CCP=critical control point

Kefir is a cultured-milk beverage believed to have originated many centuries ago in the Northern Caucasus Mountains. Kefir has a uniform creamy consistency, a slightly sour taste somewhere between buttermilk and sour cream, and a mild yeasty aroma. Kefir may have small amounts of carbonation and alcohol. It can be enjoyed plain or sweetened to taste. Traditional kefir is prepared by combining fresh milk with the Kefir culture made up of yeasts and lactic acid bacteria. Kefir’s live culture has been claimed to have health benefits similar to that of yogurt.  Kefir could be made in a retail-foodservice operation as a fermented (acidified) food with a simple HACCP plan (approved by the regulatory authority).

L. Seaton cc 2010
L. Seaton cc 2010

The kefir culture is more commonly referred to as “grains” since it forms grain-like casein-polysaccharide-microorganism particles during fermentation. The exact combination of bacteria and yeasts vary between kefir cultures, and might include: Lactococcus lactis subsp. lactis, Lactococcus lactis subsp. Cremoris, Lactococcus lactis subsp. Diacetylactis, Leuconostoc mesenteroides subsp. Cremoris, Lactobacillus kefyr, Klyveromyces marxianus var. Marxianus, and Saccharomyces unisporus. To ensure consistency and sterility, commercial producers now generally use a powdered starter culture rather than grains. However, such cultures may not form grains or continue to culture indefinitely; making kefir grains the preferred choice for individuals.  Retail-foodservice operators could use either.

Commercial production starts with whole, low-fat or skim milk, adjusted for body with nonfat milk solids. The milk is pasteurized, and then heat-treated at 203°F for 10 to 15 minutes denaturing whey proteins. For unpasteurized milk the pasteurization step would be considered a CCP (critical limit 145°F for 30 minutes or 161°F immediately).  For pasteurized milk the critical limit is 145°F for 15 seconds.  If whey proteins are not denatured they can cause a gritty product.  The heat treated milk is then cooled to 64.4 to 71.6°F, 2% to 5% kefir grains or culture are added, and the mixture is incubated at 64.4 to 71.6°F for 24 hours. After that time the pH ≤ 5.0 (CCP) and the kefir grains are sieved out and the product is pasteurized, packaged, and maintained chilled (SOP ≤ 41°F) . The final kefir product can be flavored in a manner similar to yogurt, but the flavoring cannot raise the pH > 5.0.  The addition of an active culture of kefir grains would be considered an SOP under the HACCP food safety system.  As kefir ferments the pH will continue to drop from 5 to 4.  At approximately pH 4.5 the kefir will begin to set.  Therefore, drinkable kefir should be pH 4.5 – 5 and kefir cheese would be 4-4.5 pH.

Home fermentation of kefir was traditionally a mechanism to preserve milk before the advent of refrigeration. Fermented foods are generally considered to be less likely to cause foodborne illness due to the fermentation process. The competitive activity and metabolites of the culture help to – partially or completely — kill or inhibit the growth of illness-causing microorganisms. Today however, the preservation of milk is easily accomplished using pasteurization and refrigeration, leaving kefir to be enjoyed for its flavor.

Kefir is generally considered to be safe due to the lack of evidence of foodborne illness events related to it. Properly fermented kefir (pH less than 4.5) inhibits many pathogens, but not for Escherichia coli, Listeria monocytogenes, Salmonella spp., and Yersinia enterocolitica. These pathogens may grow very slowly or just survive.  Care therefore must be taken in the fermentation of kefir to prevent the access or growth of these microorganisms after the pasteurization or heating steps.

  • Use only pasteurized milk or pasteurize as part of the process (CCP).
  • Use quality kefir grains from a reputable source (SOP).
  • Because of the small risk of pathogen growth in fermented kefir, it is NOT recommended for those with weakened immune systems, e.g. pregnant women, the elderly, the very young and the chronically ill (Policy: e.g. don’t market the kefir as a health drink to cure something).
  • Pasteurization of kefir before consumption is recommended and will kill any potential pathogenic microorganisms listed above (optional).
  • Refrigeration is required after fermentation (≤ 41°F).

Pasteurizing kefir can be accomplished by heating the kefir, after the grains have been filtered off, to 161°F for 15 seconds. Place the open jars of kefir in a hot water bath. Stir the kefir while heating until a temperature of 161°F is reached, using a quality food thermometer. Hold at this temperature, making sure the temperature does not drop below 161°F for at least 15 seconds, and then remove the jars from the bath. Cool them quickly in a cool water bath. Promptly refrigerate the kefir ≤ 41°F; it may be stored for 7-10 days. Pasteurized kefir will not have the probiotic effects of a live culture, but will be safer, especially for those with weakened immune systems.


Heller, KJ. 2001. Probiotic bacteria in fermented foods: product characteristics and starter organisms. American Journal of Clinical Nutrition (73)2:374S-379s.

Gulmez, M and A Guven. 2003. Survival of Escherichia coli O157:H7, Listeria monocytogenes 4b and Yersinia enterocolitica O3 in different yogurt and kefir combinations as prefermentation contaminants. Journal of Applied Microbiology 95:631-636.

Dom’s Kefir-making in-site Website: (Accessed 15 Nov 2004).

UV light can be used to achieve an effective pasteurization of juice to meet the 5 log pathogen reduction requirement of the FDA.  The UV light is absorbed by pathogen DNA to the extent it leads to cell death.  At least one manufacturer has demonstrated effectiveness for killing both Escherichia coli 0157:H7 and Cryptosporidium parvum oocysts. At least one research paper also described a reduction in patulin (a fungal toxin) using the UV process.

  • No endorsement intended, but readers can be directed to Cidersure for more information on their equipment.
  • Illinois Institute of Technology article
  • Journal article on reduction of patulin in UV treated apple juice here

There are several solutions that work for refrigeration temperature monitoring. Here are two options that come from Thermoworks, a company in Utah.

The first product is a temperature datalogger that requires a docking station to upload measurements.  This product has been tested by the author and it works as described by the manufacturer.  First – the two probe version is recommended.  That way the operator can measure ambient (refrigeration air) temperature with one probe AND the second can be placed between product bags for product temperature.  Note that the food temperature is the critical measurement!  This solution is ~$150 for the datalogger and ~$70 for the cradle.  This datalogger has an alarm light that can be set at the critical limit or a desired temperature.  It is recommended to set this system up to take a measurement every 1-5 minutes.  For ROP in foodservice a manual inspection is required twice daily, including checking a separate thermometer placed in the refrigerator.  The operator must check to ensure the alarm light is not triggered.  After two weeks the operator downloads the data to their computer for storage.  Keep those records for 6 months then discard.

thermadata_logger_td2f_c_a  293-804_ThermaData_Cradle

The second option is identical to the above system, except that it is wifi enabled.  The author has not tested this system yet for ROP refrigeration use.  The same recommendation for the two probe version is made here.  Note that this system does not have a display.  The manufacturer states that these loggers (~$200) have a range of 300 feet (line of sight) and transmit to a USB wireless base station that is easily setup on a PC (~$250).   They also mention that alarm conditions are monitored and text messages can be sent when alarm conditions have been breached.  Since this unit has not been tested by the author, there is a question of how well the signal will transmit from inside a refrigerator.  For the twice daily verification required in the Food Code the operator still needs a separate thermometer.

thermadata_logger_wtb2f_z_a rf_thermadata_base_station_squared_a

Calibration: as with all temperature measurement devises calibration is required.  It is recommended bi-weekly that probes are placed into a properly prepared ice bath that has been verified to be at 32F.  For the USB datalogger this should be done each time the unit is downloaded to the computer.  For the wifi unit, simply test each probe using the ice bath method biweekly.

Many operators struggle to find the right solution for labeling ROP bags under the Food Code and their HACCP program.  Here a few reviews of products that have worked in our ROP lab to get you started.  (No endorsements implied, just tested).

zd500First there are two types of printing technologies for labels: direct thermal and thermal transfer. Direct thermal is the simplest.  There is no ribbon and the image is heat etched on the label.  Direct thermal media is more sensitive to light, heat and abrasion, which reduces the life of the printed material.  Thermal transfer printing uses a heated ribbon to produce durable, long-lasting images on a wide variety of materials.  The direct thermal should work well for ROP.  Labels are generally not needed for more than 30 days.

The printer that’s been tested is a Zebra brand.  Others may be equally suited.  They simply were not tested.  The Zebra model ZD 500 has several features that make it suitable for ROP at retail and food service.  Read the technical specifications here. The cost is approximately 575 dollars.  It does have both direct thermal and ribbon thermal.  This model would work best to integrate with many electronic HACCP systems in current use.  Cheaper models could suffice, but will have less functionality.

Labels require the ability to withstand cooking temperatures followed by cooling and refrigeration (or freezing) temperatures.  They must also withstand water submersion.  This latter requirement requires a polypro (plastic) versus paper label.  The Zebra label Polypro 4000 D has worked well in the test lab.  It has an all temperature adhesive that sticks well during cooking, cooling, and refrigeration.  We’ve not tested the labels under rough handling such as in tumble or jet chillers.  Unfortunately there are only a few sizes available.  The good news is that purchased in bulk each label might be less than 1-3 cents.  Fortunately, because labels are consumables, many vendors will send you samples to work with.  That way you can ensure they work before you buy them.

November 2014