Cold vs. Iced: Using Sensory Analysis to Test the Claim that Cold Brew is Sweeter and Less Acidic | 25, Issue 19
Lead author Dr. MACKENZIE BATALI shares the results of a controlled and systematic study exploring the impact of the cold brew process on the sensory profile of the beverage it produces, recently published in MDPI’s open access journal, Foods.
It was an unreasonably warm day in the autumn when I visited a local diner; I flipped to the back of the menu to look for something to cool me down, and that’s when I saw it—the increasingly ubiquitous cold brew. “New!” the glossy page beckoned, “sweet and smooth”—words which triggered a thought of something delightfully refreshing.
How much thought went into these words, I wondered—was it merely marketing, an affinity for alliteration? Or was it something deeper, an understanding of the importance of sensory attributes and their effect on the consumer? For those in the specialty coffee community, we know that what’s in the cup—the flavor, body, aroma, etc.—deeply impacts consumer behavior (i.e., purchase intent, customer loyalty); therefore, understandably, the management of sensory attributes (i.e., enhancing desirable attributes and/or suppressing undesirable attributes) is increasingly important in a growing and competitive market—and cold brew’s share of that market is increasing.
According to the National Coffee Association’s 2022 National Coffee Data Trends Specialty Coffee Report, 7% of American (ages 18+) drank a cold brew coffee in the past day and 16% had one in the past week— and that was in January! Additionally, cold brewing was found to be the third most common preparation method among past-day specialty drinkers (20%), according to the same report, produced in collaboration with the Specialty Coffee Association.
Actors in the coffee value chain looking to capitalize on cold brew’s popularity may themselves be familiar with the perceived sensory differences between hot brewed and cold brewed coffees—a familiarity gained through personal and professional experience. Some of that experience may even be tied to non-coffee activities: did you ever make tie-dye at camp or school, maybe at home on summer break? Did you notice that hot water makes for more saturated color? This is because hot water extracts soluble chemicals more quickly. We even have a term for this in coffee, total dissolved solids.
Even with some experience and basic knowledge, business decisions are best supported by substantiated qualitative analysis (and less on camp crafts) required when you want to ask (and answer) the question, what is the difference?
Enter researcher Dr. Mackenzie Batali and the team at the UC Davis Coffee Center: Beginning in 2021, they embarked on a thorough exploration of cold brewed extraction techniques including temperature during brewing, brewing time, and extraction percentage and concentration; throughout, using various specialty coffee and roasts, the team documented the comparative impact of cold brew on sensory and chemical attributes.
Mary Basco
Research and Knowledge Development Programs Manager
As cold brew coffee has grown in popularity in recent years, so too has the number of strong opinions about the brewing method’s impact (positive or negative) on the flavors in the resulting cup—but they’d never been systematically studied.
Cold brew coffee is not the only form of chilled coffee consumed; its brewing method is different compared to other iced coffees. Whereas iced coffee and espresso drinks are often brewed identically to their hot counterparts before chilling with ice, cold brew is a long, slow, usually immersion-style brew with room temperature or chilled water for anywhere from 4 to 24 hours. The common belief is that this brew style produces a sweeter, smoother, and less acidic coffee compared to a coffee that is brewed hot and then chilled. Although informed by coffee professionals’ extensive personal experience, these claims about the sensory differences between cold brewed coffee and iced coffee are not the result of a controlled and systematic study.
As any coffee aficionado knows, there are a variety of factors in brewing that can contribute to the sensory profile of the cup. The UC Davis Coffee Center has previously reported on the sensory differences that come from the “strength” of the brew (the total dissolved solids, or TDS), as well as the extraction yield.[1] Both of these physical measurements will be impacted by the water temperature: hotter water extracts more material, more quickly. This is, in theory, accounted for by the longer brew time of cold brew coffee, but is—in practice—unlikely to be a 1:1 comparison. Another factor that impacts a cup’s sensory profile is filtration: for example, an immersion cold brew in a reusable mesh bag, a paper-filtered iced drip coffee, or an iced Americano, even at the same strength, would likely have different sensory profiles. For example, a 2013 study by Gloess et al., in their previous work, showed that espresso-style extractions were more roasted and bitter, whereas filter coffee and French press were more balanced and sweet.[2] And, of course, the temperature at which a beverage is consumed makes a difference in the drinkers’ sensory perception. This is not so relevant when comparing cold brewed coffee to iced coffee, but a drinker will always perceive differences in the sensorial profile of cold coffee and hot coffee, regardless of how similar every other factor is.
With all this in mind, the team at UC Davis, with the support of the Coffee Science Foundation and Toddy, designed a study to explore common beliefs about cold brew through a systematic comparison of cold brewed coffee to hot brewed coffee, removing as many variables as possible to compare only the impact of brew temperature on the sensory attributes of the coffee. We also wanted to understand if there was an interaction between the brew temperature and the coffee brewed. In other words, would coffees with a particular roast level, processing method, or origin be particularly impacted by (or more suited to) a cold or hot brew method?
Attempting to Isolate Brewing Variables
To conduct this study, we used what is known as a “three-level full factorial design,” which means we chose three initial factors—in this case, three different green coffees—which were then explored across two more “factors” (in this case, roast level and then brewing temperature) for a total of 27 samples. The three green coffees we chose were a honey process from El Salvador (ELS), a washed coffee from Ethiopia (ETH), and a wet-hulled coffee from Sumatra (SUM), all of which had additional associated extrinsic attributes (like region or certification).[3] These were roasted to three different roast levels (light, medium, and dark),[4] for a total of nine different roasted coffees which were then brewed at three different brewing temperatures: refrigerated (4°C/39.2°F), ambient (22°C/71.6°F), or hot (92°C/197.5°F). We used a full immersion brewing method for all the samples, applying a 5:1 brew ratio in Toddy Cupping Kits filtered with Toddy brand paper filters. Each sample was brewed to equilibrium extraction, or allowed to continue brewing until the total dissolved solids stabilized and no longer increased (36 hours for 4°C, 12 hours for 22°C, and 1 hour for 92°C). Samples were then diluted to a TDS of 2%, and chilled until the time of serving.
By designing the study to have samples brew to equilibrium instead of a fixed brew time, we wanted to remove variability in the samples’ extraction yield, as different origins, processing methods, and roast levels may have slightly different rates of extraction. In diluting each sample to a fixed TDS, we wanted to remove variability in each samples’ strength. Ultimately, our goal was to ensure that any differences we found could be attributed to the brewing temperature alone (or interaction between a coffee’s origin or roast level and the brewing temperature), even if the methodology used did not perfectly reflect standard brewing practices. We know, based on previous studies, that TDS and percent extraction strongly impact sensorial properties of coffee,[5],[6] so if we wanted to make rigorously supported claims about the impact of temperature, our experimental design would need to take this into consideration.
As we were planning the sensory portion of the study, the COVID-19 pandemic hit, causing us to entirely rethink our sensory experimental design: gathering tasters in a sensory assessment lab was simply not possible! We identified a way to conduct the study entirely remotely, using a modified descriptive analysis methodology where panelists local to UC Davis were delivered samples and sensory references before attending Zoom sessions together to develop the sensory ballot and discuss. Sensory descriptive analysis functions somewhat differently than traditional coffee cupping, as it is a much more objective observation of the attributes in the cup, with no judgement of preference. Panelists were trained on the 27 attributes, and then asked to evaluate each sample on intensity only, on a scale from 0 to 100. This again was done remotely, with panelists receiving samples delivered to their home and consuming the samples chilled (to control for the impact of consumption temperature on sensorial properties), before logging into an online survey where they noted their evaluation of each sample. In the end, we came away with 1,134 sample evaluations: each of the 14 trained panelists evaluated each of the 27 samples three times! Figure 1 shows a full schematic of the brewing and evaluation process.
Figure 1. Full schematic of the brewing process used to produce the samples for evaluation.
Figure 2. Sensory differences by temperature. Hot brewed coffee was perceived to be more "bitter," "sour," and "rubbery"; cold brewed coffee, more "floral."
What We Learned
While the goal of brewing to equilibrium was to eliminate the impact of extraction yield, it became immediately obvious that it would not be that simple! Of the three brew temperatures, the hot brewed coffee had the highest equilibrium TDS (4.27%) and extraction yield (or EY, 22.28%). Samples brewed at an ambient water temperature had slightly lower equilibrium TDS (3.98%) and EY (20.74%), and those samples brewed with refrigerated temperature water had the lowest (equilibrium TDS of 3.76%, EY of 19.55%). We also found that the roast level impacted equilibrium TDS and EY too: the light roast coffees also had a slightly higher equilibrium TDS and EY compared to the over roast levels, consistently across all brew temperatures. This suggests that there might be some molecular components of coffee that do not extract at low temperatures regardless of extraction time, but further chemical analysis would be required if we wanted to try to identify what these components are. We also found that—just as coffee professionals’ experience suggested—the cold brewed coffee consistently had a slightly higher pH, or was less acidic, than the samples brewed with hot water: this finding would also benefit from further chemical exploration into the composition of coffee at different extraction temperatures.
In reviewing the sensory panelists’ evaluations, we identified four key differences between cold and hot brewed coffee, regardless of the coffee brewed: hot brewed coffee was perceived as more “bitter,” “sour,” and “rubbery”; cold brewed coffee was perceived as more “floral” (Figure 2). The panelists’ increased perception of sourness in the hot brewed coffee also correlates with the pH measurements we took, which found a lower pH in hot brewed coffee and a higher pH in cold brewed coffee. Panelists didn’t seem to notice differences between the coffees brewed with refrigerated and ambient temperature water—good news for anyone who wants to maximize their brewing process for efficiency! (Remember, ambient temperature water took only 12 hours to reach equilibrium TDS compared to the refrigerated water’s 36 hours.)
Our sensory results also confirmed that the three different coffees and roast levels were each impacted differently by brewing temperature (see Figure 3, differences by temperature for different roast levels). All three green coffee samples, roasted to a light profile, did not have any additional differences between cold and hot brewed coffee beyond those identified based on brew water temperature alone (above). However, at medium roast level, panelists identified more “burnt” and “whiskey” flavors in the hot brew. At dark roast level, panelists identified more “woody” flavor in the samples brewed with refrigerated and ambient temperature water. Across the green coffee’s origins, we found other trends too: panelists identified more “burnt” flavor in the Ethiopian samples, brewed hot, and “fruity” flavor, brewed cold; more “woody” flavor in cold brewed samples from El Salvador; and more “smoky” flavor in hot brewed samples from Sumatra.
Figure 3. Sensory differences by temperature for different roast levels. Those samples roasted to a "light" profile did not have any additional differences between cold and hot brewed coffee, but at "medium" and "dark" roast levels, panelists identified additional differences, shown here.
Ultimately, however, although we found sensorial differences between the samples when comparing the temperature of the brewing water, these were much smaller in magnitude than those found when we reviewed the same results for differences across roast level or origin. This is an important finding: a coffee with more “rubber,” “bitter,” and “sour” flavor profile might not be enjoyed by coffee consumers, but even a noticeable difference in the presence of these flavors might not be substantial enough to change a consumer’s preference for a cold brewed coffee instead of a hot brewed iced coffee. Further testing with consumers would confirm or deny this, but we would conclude based on these results that the quality of the coffee brewed makes a much bigger difference than do the subtle differences from the brew temperature itself. Additionally, as we were unable to have all the samples reach the same equilibrium TDS, it’s possible that the differences perceived by the panelists have been more related to the extraction yield than to the brewing water temperature. We’re planning more studies to investigate this variable further by specifically exploring the relationship between brew water temperature and brew time required to reach equilibrium extraction.
We were particularly interested in the finding that the roast level and origin of a coffee impacted how its sensory profile was perceived when brewed with water at different temperatures. From a practical standpoint, this suggests that some coffees are “better” suited for cold brew and some are better suited for hot brew, and that, while it’s ultimately up to personal preference, it may be valuable to experiment with which coffees are best suited for which brewing methods by exploring how different coffees’ sensory attributes change with brewing temperature.
And that’s the point: whether the cold-brewed method is better (or worse) than the hot-brewed-then-iced method will always come down to the preference of the person brewing (or drinking) the coffee. There is some validity to claims that cold brew is less acidic, and although the temperature of the brew water does make a small but noticeable difference when consumed at the same temperature and TDS, the coffee you choose to brew will ultimately make more of an impact on the flavor profile in the cup. Regardless of where you fall on the cold vs. iced spectrum in terms of your own preference, there’s no doubt that systematic research conducted on these kinds of questions shed more light on how many variables there are at play in brewing and how an adjustment to just one of those variables can impact the sensory profile of the coffee in your cup. ◇
Now a food scientist at Compound Foods in San Francisco, Dr. MACKENZIE BATALI completed a PhD in Food Science at the University of California Davis Coffee Center, studying the impact of brewing parameters on the sensory and chemical profiles of brewed coffee. The research summarized here, supported by the Coffee Science Foundation and Toddy LLC, was completed as a part of their postdoctoral work, also at UC Davis, from January to December 2021.
A full academic paper on this research, “Sensory Analysis of Full Immersion Coffee: Cold Brew is More Floral, and Less Bitter, Sour, and Rubbery than Hot Brew,” was published in Foods 2022, and was co-authored by LIK XIAN LIM, JIEXIN LIANG, SARA E. YEAGER, ASHLEY N. THOMPSON, JULIET HAN, and Professors WILLIAM D. RISTENPART and JEAN-XAVIER GUINARD. It’s available to read in full, as an open access publication, at doi.org/10.3390/foods11162440.
To learn more about the Coffee Science Foundation’s project “Towards a Deeper Understanding of Cold Brew Coffee,” or to read other related academic papers, visit coffeescience.foundation/understanding-cold-brew.
References
[1] Mackenzie E. Batali, William D. Ristenpart, and Jean-Xavier Guinard, “Brew Temperature, at Fixed Brew Strength and Extraction, Has Little Impact on the Sensory Profile of Drip Brew Coffee”, Sci. Rep. 10, no. 1 (2020): 16450, https://doi.org/10.1038/s41598-020-73341-4.
[2] Alexia N. Gloess, Barbara Schönbächler, Babette Klopprogge, Lucio D’Ambrosio, Karin Chatelain, Annette Bongartz, André Strittmatter, Markus Rast, and Chahan Yeretzian, “Comparison of Nine Common Coffee Extraction Methods: Instrumental and Sensory Analysis,” Eur. Food Res. Technol 236, no. 4 (2013): 607–627, https://doi.org/10.1007/s00217-013-1917-x.
[3] ELS: El Salvador Las Ranas honey process; ETH: Ethiopia Guji washed organic; SUM: Sumatra Fair Trade Organic wet-hulled.
[4] Light: Agtron Gourmet score 58; Medium: Agtron Gourmet score 48; Dark: Agtron Gourmet score 38.
[5] Mackenzie E. Batali, William D. Ristenpart, and Jean-Xavier Guinard, “Brew Temperature, at Fixed Brew Strength and Extraction, Has Little Impact on the Sensory Profile of Drip Brew Coffee”, Sci. Rep. 10, no. 1 (2020): 16450, https://doi.org/10.1038/s41598-020-73341-4.
[6] Scott C. Frost, William D. Ristenpart, and Jean-Xavier Guinard, “Effects of Brew Strength, Brew Yield, and Roast on the Sensory Quality of Drip Brewed Coffee,” J. Food Sci. 85, no. 8 (2020): 2530–2543, https://doi.org/10.1111/1750-3841.15326.
We hope you are as excited as we are about the release of 25, Issue 19. This issue of 25 is made possible with the contributions of specialty coffee businesses who support the activities of the Specialty Coffee Association through its underwriting and sponsorship programs. Learn more about our underwriters here.