Fusel Oil Recycle—A Silent, Odorous Killer

A normal part of fermentation, fusel oil concentration must be avoided
By Dennis Bayrock | May 10, 2012

There’s a silent killer lurking in your ethanol plant. You won’t see or hear it coming—you’ll have to sniff it out. Fusel compounds are oily, odorous byproducts of the ethanol production process that slow fermentation and kill yeast. If allowed to recycle unchecked, fusel compounds can even render a fermenter completely sterile, halting production altogether. Fusel compounds are also found in emissions in CO2 scrubbers, dryers and thermal oxidizers (areas under increased scrutiny by the U.S. EPA).

Fusel oil production is a natural and normal part of fermentation by Saccharomyces cerevisiae. A 100 MMgy ethanol plant produces nearly 40,000 gallons of fusel oil a year, so dealing with it properly is critical. Technically, fusel oil is not a single substance, but rather a mixture of volatile organic acids, higher alcohols, aldehydes, ketones, fatty acids and esters. When concentrated, this mixture has an oily consistency and potent odor. The fusel compounds of primary interest to an ethanol producer include amyl alcohols (isomers of C5H12O such as isoamyl alcohol), 1- and 2-propanols, butanols (such as n butanol, isobutanol) and other volatile compounds.

Fusel compounds pose a significant risk to ethanol yield. If not properly managed, they can become part of the process water and recycle to the front end of the plant. Both ethanol and fusel compounds are toxic to yeast growth and fermentation, but fusel compounds are 10 to 15 times more toxic to yeast than ethanol. Removing fusel compounds properly is accomplished using a rectifier distillation column, which adds capital, energy and operating cost. By adjusting the pressure, flow and operating temperature in the rectifier column, the fusel compounds are removed from the ethanol via draw points on the side of the column and a fusel draw pump.

Why do yeasts make fusel compounds at all? The answer lies deep in their metabolism. As yeasts ferment glucose into ethanol, another metabolic intermediate is produced—a cofactor molecule called NAD. As ethanol production increases, so does the NAD concentration. This creates a serious problem for the yeast cell, as it needs to regenerate the NAD back to its NADH form so that it can be used again. One way it can regenerate the NAD is with the production of glycerol. This regeneration route explains why yeasts must produce glycerol in addition to ethanol, and why production levels of glycerol increase with increasing ethanol production.

Yeast cells can also regenerate the NAD with a series of metabolic reactions included in the Ehrlich pathway. The Ehrlich pathway regenerates the NAD for the yeast by breaking down individual amino acids inside the yeast cell. This is responsible for the majority of fusel compounds produced by yeast.

When more than one fusel compound is present, the inhibitory effects on yeast are synergistic. For example, ethanol at a 5 percent weight/volume concentration begins to slow the metabolism of most yeast. With fusel compounds, a 0.5 percent weight/volume concentration can have the same effect. This danger has not only been confirmed in literature, but within the last two years, a growing number of fermenters at ethanol plants in North America have been rendered sterile.

How does one determine if fusels are a problem? Let’s walk through what happened during an emergency forensic audit at one ethanol plant earlier this year.

Fusel Oil Case Study
For the plant, the first sign of the emergency was the 1 percent weight/volume drop in absolute ethanol yield in the fermenters from baseline levels. Additional symptoms were discovered on-site:
• All the fermenters stalled at 40 hours, regardless of any interventions tried by the plant prior to the audit.
• The plant was inconsistently passing 2 to 4 percent weight/volume total sugars from the beerwell to the beer column.
• The lactic and acetic acid deltas across all fermentations were at a maximum of 0.1 percent weight/volume, indicating that the problem was not due to a bacterial contamination.
• The alpha amylase and glucoamylase conversion rates were still proceeding from DP4 to DP1 at the plant at baseline levels. This eliminated the possibility of incorrect enzyme dosing or process conditions affecting the enzymes.
• The glycerol concentration decreased below 1.7 percent weight/volume (plant baseline). This, together with the lower ethanol and acetic acid levels, indicated that the yeasts were either metabolically inhibited (assuming correct crop of yeast entering the fermenters), or  not enough yeast was entering the fer menters.
• The yeast crop leaving the propagators to the fermenters was below 200 million cells per milliliter (ranging from 100 to 150). 

Out of all the possible diagnoses these symptoms could indicate, experience and intimate knowledge of yeast narrowed it down to three possibilities. First, the growth and ethanol production rate of the yeast could be chemically inhibited, caused by fusels or any number of process chemicals. Second, the yeast perhaps was nutritionally limited. Or, third, insufficient yeast may have been leaving the propagator for the fermenters, causing a slowdown in overall fermentation rate.

To isolate the problem, we pulled multiple samples for microbiological analysis. No significant level of bacterial infection was found, which was consistent with current plant data on lactic and acetic acid levels. We also immediately had samples analyzed for mycotoxin, sodium and minerals. The results indicated that these chemicals were at normal levels and could not be the source of the yeast inhibition.

Next, we focused on learning the actual layout and operation of the plant, including any changes made since construction. An examination of the plant’s fermenter/propagator fill sequence indicated that no changes were made when the emergency started. Similarly, the cleaning in place (CIP) system and sequence were free of any changes or problems.

The breakthrough came when we again pulled mash samples at various locations and gave them the sniff test. The slurry sample had a strong varnish-like smell. We suspected a fusel oil problem, as higher temperatures during slurry would tend to flash off any ammonia, CO2, ethanol, hydrogen sulfide, chlorine and sulfites. (Many fusels boil only at temperatures greater than 100 Celsius.) The samples were left overnight and reexamined the following day. The smell remained, pointing again to fusel oils.

We examined (and smelled) all of the lines leading to the slurry tank. The source of the smell came from the side-stripper and rectifier columns. Further examination of the plumbing around the rectifier revealed that the fusel draw pump had broken down and needed to be replaced.

By this time, all of the liquid from the slurry to the distillation was contaminated with fusels. There are no known methods to remove fusels other than distillation, so we had to be patient. Once the new fusel oil pump was in place, it took approximately one week to purge the fusels and return to normal operation.

It’s a good idea to “smell your slurry” on a daily basis. Fusel compounds can often be detected by smell when present in as little as 20 parts per million (about 0.002 percent weight/volume). If detected, pay immediate attention to the rectifier column and surrounding plumbing. For a 100 MMgy ethanol plant, a 1 percent yield decrease, as seen in this case study, results in an approximate loss of $179,167 per month at $2.15 per gallon of ethanol.

Author: Dennis Bayrock,PhD
Global Director, Fermentation Research Lactrol
Phibro Ethanol Performance Group
(651) 641-2826

The claims and statements made in this article belong exclusively to the author(s) and do not necessarily reflect the views of Ethanol Producer Magazine or its advertisers. All questions pertaining to this article should be directed to the author(s).