Adventures in Coffee Tasting

Scott Westfall’s interest in chemical engineering began with a question: “Why would THAT work?"

As a Nestle R&D summer intern, he couldn't get the heat exchanger to make hot water. “Turn the flow up,” the supervisor answered.

Westfall found this counter-intuitive, and asked why. “Turbulence increases heat transfer,”  his boss said simply, and walked away.

It was ChemE 101, but it intrigued him. Following his BS degree in biochemistry at Ohio State, Westfall followed up with a 1984 MS in chemical engineering at Ohio State before resuming work at Nestle.

His Nestle career focused on instant coffee, from serving as a “taster,” to developing products and designing more efficient equipment, processes, and operations. He developed a novel aroma recovery process and an encapsulation process, and designed an innovative calorimeter and a scraped-surface heat exchanger. His job took him to Mexico, Malaysia, India, Thailand, the Philippines, and Switzerland. As principal engineer, he led a team of six engineers and managed a capital budget >$10M before retiring in 2013.

Birth of a habit

The story Westfall tells today about his work with coffee is a fascinating one, but it all started long ago. Legend has it that coffee-drinking started in Ethiopia when a shepherd saw that his goats became livelier after eating coffee cherries. Green coffee doesn’t taste very good, so when people began frying it, it made all the difference. “Something magical happens during the roasting process,” Westfall said. “It’s like a little reactor inside the bean.”

There are two types of coffee: Arabica and Robusta. Arabica, named after the Arabian peninsula from which it was introduced to the world, is now grown in Columbia, CentralAmerica and East Africa. Robusta originated in central Africa but is chiefly now grown in SE Asia. It's a smaller bean, has twice the caffeine of Arabica, is less expensive and more disease resistant, and has a stronger flavor some people describe as "rubbery."  

Both coffees share that all-important ingredient, however: caffeine. By the 15th century, coffee was being consumed as a beverage, quickly becoming habit-forming among its users and gradually spreading worldwide.

A matter of taste

People aren't born liking coffee. It's an acquired taste resulting from acclimatization. In some countries, the beans are roasted in butter due to a lack of roasters, but the butter is often rancid—an unpleasant taste to most Westerners.

In countries that only have access to lower-quality coffees, people come to prefer a certain note that, to an American, tastes like stale, overheated coffee.

These regional taste differences made overseas work challenging. To learn about regional coffee preferences and the associated terminology, Westfall and his team visited local coffee shops and participated in focus groups or in-home tastings, where they also observed preparation.

Determining the "taste" of coffee is not easy. There is no chemical model for coffee; one can only match some of the chemistry. Researchers have been able to determine that the bitter taste of coffee is one-third lactones and diketo piperazine and one-third caffeine, but the other third remains unknown.

To complicate matters, unlike color, which is linear, taste and smell intertwine to affect one another. When two tastes combine, a third taste is produced, and aroma also affects taste, as illustrated by differing scores when tasters wear nose plugs. Aroma itself is complex, with 1100 compounds in coffee aroma. Which ones are responsible for the actual aroma? About 70 of them, it turns out.

When the “ideal” taste has been identified, the R&D team uses the coffee to perform a principal component analysis, identifying discriminating sensory attributes. The next challenge is to get these attributes into the product in the right ratios in order to bring out the characteristics important for that market.

“Producing a ‘good’ instant coffee is a matter of trade-offs,” Westfall said. “The product has to be cost-efficient while having the desired taste, aroma, freshness, and ease of preparation."

“I love going into a factory and solving problems.” – Scott Westfall, '84

To supplant the Philippine market, the team was charged with producing Philippine-style coffee within the Thai market using only locally sourced coffee due to import restriction.

In Mexico where demand for Nestlé’s product outstripped the local supply of coffee, it was a challenge to get enough coffee beans, again due to import limits. Nestle imported all the coffee they could, but had to supplement this with local poorer quality beans. From this, Westfall's team had to create a product that would match consumer expectations.

The R&D team did not just work on taste attributes, but caffeine levels. When the Thai government banned amphetamines, truck drivers wanted something else to keep them awake on the road. Nestle filled the gap with “Gorilla blend,” a robusta-based coffee cranked-up through a flavor recovery process. Thai truck drivers loved it, but Americans would have found the taste hard to stomach.  

Nestle also developed caffeine-free coffee. One variety was not pursued because it included GMOs— not popular with the public. This particular coffee would also have been hard to grow. Caffeine is a natural pesticide, so the coffee would have needed the addition of pesticides.

A cup o' trouble

The production of instant coffee has its share of equipment troubles. Several times, Westfall was rushed to the site of an “equipment failure,” only to find operator error.

After having spent $10,000 on a plane ticket to the site of one such emergency, he found that the factory had recently installed a sample recycle port. The operators were leaving the valve to this port open at all times. The problem was that when draining the weigh bowl, the coffee would aspirate air, creating mounds of foam spilling from the tanks that plant operators then had to walk through.

Standing one's ground

Engineers are trained to determine best outcomes, but that can get tricky when discovering an

improvement that contradicts traditional wisdom, especially when one’s company has just invested $100 million dollars in doing things the old way. “I discovered that Method A was actually the worst way, and that we should be doing Method B,” Westfall said, recalling one such incident. "I nearly lost my job over it."

“No organization survives bad decisions." Westfall explained when asked why he kept speaking up, despite the risk. "If you don’t challenge what appears to you as a bad decision, then you are complicit in the failure.”

In this case, Westfall won his bosses over by providing evidence that his way of handling materials was both cheaper and easier, and it was implemented in 40 factories.

Later, Westfall told his team, “If your boss is wrong, say so and put forth a strong case. If your boss objects, say so again. But if your boss still objects, then the boss is right, and it’s time to write a resume.”

A tasting session described

Tasters underwent initial and ongoing training. A typical tasting session: Each taster sits in an isolated booth. The lighting is controlled to a deep red so that any visual differences in the cup aren't apparent. The taster is provided with a hot water pot and preweighed coffee samples in white porcelain cups. The taster pours the proper amount of water into a graduated cylinder and then pours this into the cup. The taster sniffs the cup and rates various aroma qualities using software with virtual “sliders” to select a value between 0 and 10 for each attribute.

Then the tasting begins. The tasters, using a standard tasting spoon, slurped the coffee. This is a very noisy slurp designed to pull aroma and flavor into the retronasal passages. The taster then spits the coffee into a "dentist bowl" and scores the sample.

Talking is prohibited. But being human, when tasting something like one of the generic, low-cost overseas products, you might definitely hear something like, "Oh my God!” at which point the sensory expert would remind the tasters not to talk. 

After some tasting sessions, the tasters would gather and the composite score was revealed. Tasters were then given their individual scores. If a taster disagreed with the panel, they could discuss why. The group tried to achieve uniformity, but too much agreement can lead to overly conservative scoring. If everyone on the panel gives the medium score of 5 for each attribute, you have absolutely no sample differentiation.

At times there were competitions among the panelists as a means of honing their skills and determining the better tasters. Just like James Bond can identify Château Lafite Rothschild Cabernet Sauvignon with a sip, the tasters were to identify coffee based on type and origin. 

Skilled tasters earn the title of “Master Taster” and are renowned for their ability to identify subtle taste differences, even when the evidence suggests otherwise. Westfall recalled a master taster named José Jimenez who was once offered scotch from a Chivas bottle. Jiminez said that it tasted like McMasters scotch. The host then confessed that it was, in fact, McMasters scotch in the bottle– not Chivas, as one would assume.