Coffee Freshness: A Test of Ideal Shelf Life

This article is a precursor to a larger study that is going on at CSEC, looking at ageing chemistry occurring to the coffee during storage over time.

We’ve seen a lot of opinions on how long is best to “rest” roasted coffee before drinking. Even within our own company, we’ve had a variety of opinions, some shaped by tradition, some by world barista competition winners, and some by personal preference.

It’s confusing.

So, we decided to put some scientific method around the issue. Herein (and in successive posts as we get new information), we seek to answer the question:

How long should I age my coffee before drinking it?

In short, we found that:

  • coffee in standard packaging (i.e. not gas-flushed) is better to drink within 1-3 weeks from roast (ideally around 1.5 weeks).
  • coffee in gas-flushed packaging is at its prime after 2 weeks of shelf time.   

coffee freshness graph

Figure 1: Recommended ageing period before consumption. We recommend using non-gas-flushed coffees (i.e. standard packaging) any time after 1 week of ageing. Gas-flushed coffees are ideal black from 2 weeks onwards. An exception to this rule is for coffees served with milk – we recommend using gas-flushed coffees from 1-week ageing period onwards. The harshness of the coffee at one week ‘cuts through’ with milk, significantly out-performing its non-flushed cousin.

How did we determine this?

First things first: What is Ageing?

Ageing refers to the development of coffee flavours over time, typically with positive connotation. Staling refers to the same thing with negative connotation. Whether you refer to the development as staling or aging, detectable taste changes occur to roasted coffee over time.

Changes to a coffees taste result from two different natural phenomena:

  • Chemical transportation
  • Chemical transformation

Chemical transportation refers to the movement of compounds from the coffee to the surrounding atmosphere via diffusion and mass transfer. This encompasses the movement of CO2, volatiles, water and vapor-phase equilibrium products of flavonoids, fats and other solid/liquid compounds within the coffee matrix to its surrounding atmosphere. Similarly, chemical transportation includes the movement of CO2, volatiles, water, air and all other chemicals BACK INTO the matrix. These phenomena are illustrated in the cartoon below.

Figure 2: Coffee flavours and other chemicals formed from roasting physically move from the coffee to the gas phase and vice-versa.

The second refers to all chemical reactions that occur over time, which includes vapor, liquid and solid phase reactions. These can include oxidation, reduction, polymerization (including dimerization), hydrolysis and others. The reactions change the chemical makeup of coffee over time, which can significantly change one’s perception of taste and aroma.

coffee grind fresh

Figure 3: Coffee also undergoes chemical reactions in both phases. The products of these reactions then continue to move through both phases, changing the taste over time.

It’s a complex business. The complexity is further increased when you take into varying account coffee origins, roast degree, water activity (pre- and post-roast) and atmospheric conditions. Accounting for every mechanism and variable that changes taste is difficult – although, we’re working on it! In the meantime, we have done some simplified studies around the ideal time post-roast to enjoy your coffee.

What did we do? (How did we test this?)

We planned our experiments to observe quantifiable and qualitative changes to three different coffees over time: a light roast blend, a medium roast blend, and a dark roast blend. These coffees were all roasted on the same day with the same environmental background conditions. We observed the difference between gas flushed** (2 parts per hundred [pph] oxygen) and non-gas flushed products (21pph). The coffees were tasted at increments of 1,2,4 and 8 weeks of shelf life.

(** – gas flushing refers to the process of adding nitrogen during packing to delay oxidation and extend shelf life)

  • We tested the coffees with five trained coffee tasters (including a certified Q-grader).
  • We took into account palate variability and recorded this in our results (i.e. differences in perceived quality).
  • We scored the coffees out of 5 in categories of sweetness, mouthfeel, milk cut-through (yes, we tested in both milk and black), and balance.
  • We also mapped the observed flavour of the coffees over the course of time.
  • Ambient temperature was not monitored or controlled in this study (Phase 2 will look at this) – however our storeroom temp is typically in the range 18C – 26C
  • Other factors that were not controlled in this test were relative humidity during roasting, holdup time in packing and ambient pressure (Phase 2)

As we all know, results from scientific work are relative. Numbers mean nothing unless there is a standardized reference point on which to compare them. So which standard do we use the calibrate our taste buds? Well, in this study we didn’t know which roast ‘shelf life’ was appropriate for a standard across the weeks of testing (though we do now! – begin phase II!). So, we reported our results as flushed vs. non-flushed. In other words, we tracked the quality of non-flushed coffee as it improved and deteriorated relative to gas-flushed coffee. We didn’t assume the gas flushed was a constant; reactions still occur under these conditions. However, what was important to observe was where the ideal quality crossover point existed (i.e. when gas flushed became superior to non-flushed), and when the coffees no longer became acceptable to drink.

What did we find?

We found that non-flushed coffee is much better drunk within 3 weeks post-roast (ideally around 1.5 weeks). We also found that gas-flushed coffees are at their prime after 2 weeks of shelf time.

We have summarized the results in a normalized score vs. weeks post-roast in the figure below.


Figure 4: Normalized quantitative results of the coffee freshness studies as a score out of 10. The error bars on the vertical axis indicate variation between tester scores, where error bars on the horizontal axis indicate average variances between light roast and dark roast (centered around ‘medium’ roast). Darker roasts tended to peak and decline earlier (thus the left-bound error bar on each point of data), which is comparable to what we have seen in the literature.

At week one, we found that all non-flushed coffees were complex and balanced (stored at 21°C in a store-room). Milk cut-through was good. We found that the flushed coffee was rather harsh when consumed ‘black’ within two weeks, though milk cut-through was noticeably better at this stage. In Australia, over 85% of the market drinks coffee with milk, this this may be a good age in which to use the coffee for the broader market.

At 4 weeks, the gas flushed coffee taste was softened, with more complex notes coming into better balance when black. Milk cut-through was still good. At 4 weeks, the non-flushed coffee had lost much of its complexity, becoming a less enjoyable in milk compared to its gas-flushed brother. The non-flushed coffee was still acceptable black, but we would not recommend non-flushed coffees be consumed much later than this, especially in milk.

At 8 weeks, the gas-flushed coffee was still delicious. The non-flushed coffee, on the other hand, was described as stale by at least 50% of the testers. This aligns well with what we found in the published literature [1],[2].

Conclusions: Ideal Coffee Freshness

At this stage we can confidently say that if your coffee is not gas flushed, drink it within 3 weeks! If your coffee is older than this, all is not lost: the coffee is still drinkable black up to 8 weeks. If you’re a café owner, you’d do well to use the coffee fresh (within 1-2 weeks of the roast date). This is true for light, medium and dark coffees.`

If your coffee is gas flushed, your coffee will stay fresh for over 8 weeks. At no point did the quality decrease over the tested timeframe – in fact it only ever increased (if only slightly), indicating that the coffee is likely to be good for longer than this. Milk cut-through is good throughout its lifetime, though if you want to serve it black, we’d recommend waiting until after 2 weeks to serve.

For reference, at Seven Miles, most of our blends are nitrogen gas-flushed – with some exceptions. Get in touch below if you’re after specifics.

Where to next? Now that we know our point of reference for flushed and non-flushed coffees, we’ll be delving deeper to understand ageing and freshness more, including:

  • What chemistry is going on over time? We’ll look at analyzing headspace gases to determine changes in chemical composition over time, including both degassing-rates and oxidation/reduction type reaction rates.
  • What rate does oxygen interact with the mixture? We’ll look at doing some oxygen extinction experiments
  • What difference does ambient humidity make to staling/aging? We’ll measure water activity over time post-roast
  • Finally, we’ll redo some of the taste tests over a longer period of time – using the results of this test as our reference point

Stay tuned!

If you have questions about this research feel free to drop us a line, or come along to one of our live events.

For more on the work of the Coffee Science & Education Centre, click here

[1] doi:10.1006/fstl.2000.0732: Cardelli and Labuzza, Lebensm.-Wiss. u.-Technol., 34, pp. 273-278 (2001)

[2] doi:10.1016/j.lwt.2008.01.011: Guerra et al., LWT – Food Science and Technology 41, pp. 2070 – 2078 (2008)

Adam Carr
Coffee Science and Education Centre (CSEC) Ph.D, B.E. (Chemical Engineering) Adam has worked in research for over 8 years, with an overarching focus on high temperature water technology. He has worked in Australia at the University of New South Wales, and internationally at Cornell University, the Massachusetts Institute of Technology and Aerodyne Research Inc. The applications for his research have been diverse, including advanced particle production technologies, reaction engineering, sustainable fuels processing and novel extraction technologies. Once he discovered the application of his experience to the coffee industry, there was no going back! Adam returned to Australia in 2015 to pursue his interest in designing his own coffee technologies. He created the startup Highpresso PTY LTD to provide research consulting services to the coffee industry, and build new coffee technologies. His research led him to designing a new roasting system that prevented charring of the raw beans. Now at Seven Miles, Adam uses his skills and deep love of coffee to improve our understanding of coffee science and technology through our coffee science and research center. He designs experiments around testing the latest and greatest coffee gear available to the industry, and uses his connections to develop research projects centering around brewing better coffee.