Posted: 28 Jul 15
Updated: 30 May 17
The goal of this study was to examine the impact of tamping pressure on espresso extraction. The main metric used was TDS assessed via two refractometers (VST LAB Coffee III and Atago PAL-COFFEE). Four different tamping pressures (5 kg, 10 kg, 15 kg, and 20 kg — or 12 lb, 22 lb, 32 lb, 42 lb) were assessed, with 10 shots per condition. No significant difference was found between the measurements of the VST or Atago device. Further, no significant difference in TDS (and, by extension, extraction yield) was observed due to tamping pressure. Brew weight had a significant negative correlation with TDS (p = 0.00).
Within the tamping pressure range we tested, no difference was seen in the TDS of the espresso. Further, no difference was observed in the readings from the Atago PAL-COFFEE and VST Lab Coffee III refractometers.
We have no vested interest in any of the products being used for this experiment.
Jeremy and Joe
At a barista competition several years ago, a former World Barista Champion mentioned he “barely tamped” as he did not believe tamping pressure played a significant role in an espresso’s extraction. This planted a seed for an idea we’ve been wanting to test in a controlled, systematic way for many years now. Further, before we move forward with other assessments relating to tamping, we decided to better understand the impact of some basic variables. With a setup capable of ensuring a completely level tamp, we set out to test the influence of tamping pressure on extraction.
The two main hypotheses were 1) tamping pressure would have a significant effect on TDS and 2) the Atago PAL-COFFEE and VST LAB III would give significantly different TDS readings for the same sample of coffee. Neither of these were supported by our data.
The coffee used for the experiment was MC Euro Classic Espresso Blend. Coffee was rested for 11 days before use. The same batch of roasted coffee was used for the entire experiment.
- La Marzocco (LM) GS/3 MP (mechanical paddle), single group fitted with 0.6mm restrictor and a naked portafilter with the 20g VST basket filter, set at 9 bar water pressure (verified with a Scace II); brewing temperature 92.5 degrees C regulated with a PID on board
- Water filter used for the experiment was Brita Purity C150 Quell ST
- Victoria Arduino Mythos One coffee grinder
- Two scales (Ohaus to measure the dry coffee dose; AWS to measure the beverage mass)
- Pen and paper to record values
- Pair of scissors
- Infra red thermometer
- 10 (for each condition) empty ramequin bowl for measuring the mass of the shot
- 10 (for each each condition) ceramic/glass cups for TDS measurement
- 2 X Smart Tamp tamper to standardize the pressure applied to the coffee (accuracy verified with a scale)
- Atago PAL-COFFEE refractometer
- VST LAB III 4th Generation refractometer
- Distilled water
- Alcohol pads
Experiment was carried out over 5 hours. The conditions with the grinder and espresso machine were constant for all conditions. The grinder blades were warmed up by grinding 500g of coffee beans. The room temperature was controlled with air conditioning at approximately 19-21 degrees Celsius for the duration of the experiment.
A brew ratio of 1g coffee to 2g brew weight was used (i.e., 20g dose for 40g final beverage mass). Because TDS is most strongly correlated with quantity of beverage, shots were all pulled to a consistent weight. Time to reach this weight was recorded, giving us a metric of flow rate (weight/time). The grinder was dialed in using 15 kg tamping pressure with a target of 40g final beverage mass in around 40 seconds. No grind adjustments were made for the entirety of the experiment.
Four different tamping pressures (5kg, 10kg, 15kg, and 20kg) were applied with a 58mm diameter tamper base, with 10 shots per condition.
Zeroing preparation of the refractometers was performed identically for all conditions.
- Prior to all measurements, the refractometers were set at room temperature together with the distilled water (a common practice we use).
- All refractometers were zeroed and cleaned prior to the coffee extraction.
- The water temperature for zeroing was between 15-30°C (as recommended by VST).
Filtering procedure for all coffees was carried out according to VST instructions.
- For each shot, we allowed the espresso to cool for 1 minute and the sample was stirred for 5 seconds.
- For each shot, the coffee crema was skimmed with a clean spoon prior to every reading
- For each espresso, 4ml of espresso was drawn up using a pipette and slowly dispensing it into a clean ceramic cup. The sample was then given approximately 45 seconds to cool down. Filters were not used for this experiment, as our other experiment demonstrated that they do not have a significant effect on the devices’ TDS readings for espresso.
Sampling procedure for all coffees was carried out according to VST instructions.
With a clean pipette, we drew a sample large enough to dispense into both refractometers. This sample was transferred to a clean, dry cup and allowed to cool for an additional 30 seconds (the temperature of the sample was checked with the infra-red thermometer). Another fresh pipette was then used to siphon the cooled espresso sample and transfer it to the refractometer wells. The coffee sample was allowed 25 seconds to equilibrate in the well of the refractometer before recording.
A mixed design analysis approach was used on the data from this study (using R 3.1.1). As per our usual protocols, data was analyzed in a de-identified manner (using labels matched to a key after all analyses were completed) by an individual not involved in the data collection. Ten samples per pressure condition were measured (40 total measurements for each refractometer). The between variable was device (Atago, VST Lab III) and the within variable was pressure (5 kg, 10 kg, 15 kg, 20 kg). Order of the samples being placed on the refractometers was pseudo-randomised to ensure a balanced order at each pressure level.
Data was first assessed to ensure it does not violate assumptions for an analysis of variance (ANOVA) or multiple linear regression analysis. A mixed model ANOVA was then performed, yielding no significant difference between the Atago and VST measurements (p = 0.70), no significant difference between pressure conditions (p = 0.40), and no significant interaction between device and pressure condition (p = 0.70). Extraction yield, of which several methods exist to calculate, was then used in place of TDS. As expected, no significant differences were seen between device, pressure condition, or the interaction between the two (all p = 0.60+ regardless of which formula was used). Because there were no statistical differences, TDS measurements from both devices were collapsed into a single variable for further analyses (unless otherwise specified).
Brew times to reach the goal weight (40g) were not significantly affected by tamping pressure (p = 0.30), but the impact of brew weight and brew time was significant on final TDS. Because of their significant correlation with each other (p = 0.00), before being entered into the regression model, the brew weight and time variables were centered (i.e., [value – mean]) as is standard (e.g., Draper & Smith, 1998). This allowed them to be entered into the multilevel model. The four-level model included the variables of pressure, device type, brew weight (centered), and brew time (centered). Results yielded an overall significant model (F4,75 = 4.181, p = 0.00) with an adjusted R2 of 0.1823. Of the predictor variables, the only significant predictor was that of brew weight (p = 0.00).
Lastly, extraction yield was, not surprisingly, significantly correlated with TDS. Though the extraction yield calculations differed significantly from one another, which extraction yield formula you use is somewhat irrelevant, as they yield differing results not due to quantities measured here but in the incorporation and/or handling of various assumptions/other variables (e.g., CO2).
(Raw data can be downloaded in a tab delimited text file here. As always, while we offer the data for your personal use, we kindly ask that you send a message to firstname.lastname@example.org before posting or presenting it in any public forum and attach appropriate acknowledgement.)
Overall, the VST and Atago devices performed comparably–the readings from the different devices are statistically indistinguishable. Tamping pressure, in the given range we assessed (5-20 kg), yielded no significant impact on TDS nor did it significantly impact the extraction itself (i.e., time to reach beverage weight). Our statistical model explained 18.23% of the variance in the TDS samples measured. It is clear that brew weight was, by far, the most important factor in determining the TDS of the espresso (and the only significant one). Below are the relative contributions of each variable to that explained variance:
It is possible that tamping pressure outside of the range we assessed significantly impacts espresso extraction. Additionally, it may be that other variables not quantified here were significantly affected. Our measurements captured the total dissolved solids content of the shot, as well as a rate metric (brew weight/total brew time), but it is possible that tamping pressure affects aspects of the extraction that we did not assess (e.g., characteristics of particular compound extraction such as lipids).
I agree with your study, but i think the important is cupping the samples, it’s true that they can achieve same TDS but is also important to understand wich coumpound are in, sugars can be replaced by fats, o acids coumpounds by quinics.. give same results on TDS% and Extraction yeld, but significantly have different taste..
But is courious to see that 5kg tamping is enought to leveling a compact the cake, enought to prevent channeling, that’s really interesting..
Thanks to share this test.
Thanks for the comment. I think you’re right on. This is something we have been discussing ourselves (and we mention in the conclusion). TDS is simply a measure of “how much stuff” the solution contains. At the end of the day, ideally, we also want to couple that with “what are the constituents of the stuff”? The reason we have stayed away from tasting samples is that it becomes very subjective and difficult to quantify. Even though we have many years of experience in coffee, we recognise that humans are simply bad at subjective assessments (e.g., look at all the work in the wine world by Robert Hodgson). Ideally, we’d have some sort of chromatography system to give a breakdown on the quantity of individual components.
Comparing the Impact of Tamper on TDS | Socratic Coffee
[…] to influence the latter. How hard one tamps affects the coffee bed’s bulk density. Our previous post demonstrated, within a range of 5-20 kg applied tamping force, there was no discernible difference […]
Joe and Jeremy-
Fantastic findings. Percolation is very complex and your study here clarifies an important variable–tamping. TDS, lengths of extraction, etc. are good measures of what is going on, and taste will be the final measure that certainly will prove to be the most puzzling. Bravo for not yet going there.
The field of “jamming” may offer some clues as to what is going on here and will help to mechanistically link any contemplated chromatographic analyses with why tamping affects perceived flavor, or measurable partition rates of coffee compounds into the brew water.
I’m going to guess that there are tamping thresholds that will radically affect TDS, extraction profiles etc. It is easy to imagine different types of coffee particle packing regimes: Simple close type packing -> serious particle deformation. You might want to see what 100 kg does….dose will likely need to be adjusted……
Keep up the great work!
…and you will need a hydraulic press! Heck, you may create a new “phase” of coffee matter. Seriously intense tamping, or pressing, might create a material akin to paper. It could be impregnated with neutral materials to facilitate and engineer percolation, by engendering more controllable interfaces.
Right– this sounds like the brew set-up that Captain Kirk uses on the Enterprise!!!!
Although I’m 6 years late in finding this post, I love the fact that you actually designed and conducted an objective study on tamping. As a synthetic chemist with plenty of experience doing extractions and column chromatography (related skills, I think!), I have dispensed with what are for me some of the more notional espresso making techniques, tamping being one of them, and focused on quality of the beans, amount of coffee, and dilution. Your findings nicely match my own (observational) experience. While there is a definitive joy in performing coffee-making rituals, some do seem to really be just that, and objectively make little difference to taste IMO as there are just too many other important variables. So unless you have a consistent and reproducible way to maintain the integrity of the pressed grounds once pressurized steam and water enter the extraction chamber, then it will all pressure-equalize regardless of pressure used on the dry coffee initially. I cannot control that as I have a fixed size extraction chamber, but maybe others can. What I try to be consistent on are: high quality coffee that I like; grind fineness; and ratio of coffee to water (dilution). I can get a very consistent and good tasting espresso (a subjective assessment) and I have (likely forever) stored my tamper! I really do enjoy my coffee and I really appreciate studies like this…thank you so much for sharing your results and conclusions!
Tamping pressure can play a significant role in extraction flow/time/volume, depending on the machine type (linear pump, vs manual lever, vs spring lever, vs pneumatic). Furthermore the grounds themselves change in their physical behavior when compacted, tamper pressure should not be treated as a linear value. Example. If you do not ground the beans, and tamp them (as an extreme example), they will level off in the basket and the tamp will remain at the same exact precise level no matter how many times you tamp. Grind them extremely coarse (like gravel), and they will behave similarly however due to the particles being much smaller, they do not have the same level of resistance, so some particles will give, allowing further, but slight compaction with additional tamping (or prolonged tamping). Grind them fine, and the particles have very little structural rigidity and can easily be deformed with each additional tamp (or as I said before, with time). This means influences on flow restriction is not at all on linear when measuring by tamp weight.
It’s a highly flawed test. Sorry.
Thanks for the comments! While you bring up excellent points regarding the complexities of espresso extraction as it relates to coffee bed dynamics, I’m not clear how those points then establish our experiment as “a highly flawed test”. Nowhere in our write-up did we claim our experiment to be an end-all-be-all, conclusive “all tamping is pointless”. As with any experiment in any domain (e.g., pharmaceuticals, social psychology, hydrogeology), the results are valid within the context they were collected and we attempt to be very clear about this by not overstating our conclusions. External validity—or, generalizability—is something we cannot say for certain, but we openly acknowledge limitations to our approach. We list all of our equipment and procedures, allowing the reader to determine if the results are relevant to his/her application. Essentially, given a machine like the one we used with no grind adjustments (and keeping all other variables the same), tamping from 5 to 20 kg yielded no difference in TDS/extraction yield. That is a valid result, within the parameters we lay out. As for assuming linear effects of tamping/tamp weight, our means difference testing approach (ANOVA) makes no linearity assumption. Perhaps there were qualitative differences we did not capture, but we are not sure how to quantify and relate to someone something such as “perceived acidity and mouthfeel change”. We also acknowledge that outside the range of pressures we tested (5-20 kg), it is quite possible there could be quantifiable, objective effects. Tamping, no doubt, can affect bed bulk density, which has been shown to impact overall permeability (King, 2008; Navarini et al., 2009).
As an experimentalist, when you say to me that our experiment is a highly flawed test, I’m looking for a methodological flaw that renders our results invalid (i.e., a pervasive confound). Not to say our experiment is perfect, but do you have specific issues you’re referring to or are you more generally saying that under different conditions tamping may have a significant impact? If the latter is the case, we agree with that statement—but that is much different than saying our experiment is flawed!
Thanks for your response. Calling your experiment highly flawed was off the mark. For that, I sincerely apologize. I just found in my own experiments that there is a direct relationship in the sponge quality of the coffee bed and the grind size. Thank you for your clarification and indeed I was referring to the latter.
I sincerely appreciate your level-headed response. Sometimes I get a bit overly passionate about things like this I forget that there are actual people on the other side of my keyboard. I do want to express my gratitude for all that you guys do to shed more light on the dark art we all know and love.
No problem — we really appreciate you engaging and helping us think about future efforts and ways to improve what we’re doing. One aspect of the site that we’ve been tweaking is the communication of our experiments and analyses, figuring out how best to describe them in a way that comes off as relevant to a barista/coffee shop owner/home enthusiast/etc. without over-reaching with regard to what our results truly indicate (i.e., the context within which they are valid). It’s a fine balance and our approach is constantly evolving! If you have other thoughts on this (or any other topics we’re investigating), please don’t hesitate to send us a message, leave a comment, etc.
Joe and Jeremey than you for the article and all the hard work!
Question: How does brew weight specifically affect the taste of the extracted coffee? Do you guys have articles or findings that show what the brew eight contributes to the overall taste of the coffee/
Thanks for the comment. You ask a tough question–brew weight is mostly a metric for us to compare “same-to-same”. Changing a variable (e.g., tamping pressure) might lead to changes in extraction, but the best way for us to compare the effect from one variable is to set a baseline that is, most often, by using a standard brew weight. Extraction decreases as more water passes through the coffee grounds (and/or more time elapses). This rate varies depending on factors such as water temperature, particle size and distribution, or pressure, but, given standard approaches to coffee brewing, the majority of a coffee’s solubles will be extracted within the first 30 or so seconds. Brew weight relates to “amount of water passed through the coffee” but also “efficiency of the water” in extracting solubles. Typically, higher brew weight will mean a more diluted coffee solution. However, manipulating variables may allow for a comparable brew weight which actually contains more dissolved solids. The picture is further complicated when we see that different variables affect the extraction rates of different solubles. TDS attempts to give us the percentage of the coffee solution that is dissolved solids. When using that number to calculate extraction yield, we first calculate the amount of mass we extracted (TDS x final beverage mass). You can think of a coffee solution as a ratio of coffee mass to water mass. As we increase overall beverage mass, the denominator is going to increase more rapidly than our numerator (that’s the decreasing extraction rate). Often we find that pleasant tasting coffee is brewed within a particular “efficiency” (i.e., extraction yield of 18-22%). Over- or under-extraction carry their own influences on perceived flavor. Here are some graphs I made of data from an old study titled “Extraction rates for selected components in coffee brew” (Merritt & Proctor, 1958).
Interesting results here. I think there might be a greater effect when using a simple steam-presssurized machine like my old Krups mini, where tamping pressure would have a direct effect on the back-pressure in the boiler, and thus on the temperature of the water. These machines operate with the water at or above 100ºC., and there’s a definite difference in taste (an improvement IMO) if you cut off extraction before the pressure (and temperature) reach maximum.
(Thanks to Dave S. for pointing me here.)
Again，this is Blake here ,and I m very interested about ur experience .Can i translate your article into Chinese and sharing in my account ?I think it would be great for barista in China.
Again，this is Blake here ,and I m very interested about ur experience .Can i translate your article into Chinese in my WeChat account ?I think barista here would love to read it.
Very interesting study! I chose the pace from these https://milkfrothertop.com/what-is-coffee-tamper/ and noticed that they are made of different materials. Tell me, does the material affect tampering?
Why didn’t you try pressures less than 5kg, to try to determine the threshold at which increasing pressure stops changing the results? Very odd.
There are many different pressures we would have liked to test. The pressures used for the study were dictated by the Smart Tamp, which had 5 kg as its lowest pressure.
Jeremy and Joe,
I am just getting into espresso and am working on getting dialed in.
Thanks for your excellent work and write up.
I enjoy the process and the product, your work helped me take a step back and enjoy both a little more.
Of special note is how well you handled negative comments.
People well skilled in physical and social science are uncommon.
I appreciate you and your work.
We really appreciate your comment! We love coffee, science, and the coffee community, so Socratic Coffee is a natural manifestation of those interests coming together. Our goal is simply to supply others with the most honest, best quality data and analysis we can given our hobby lab setup. And we hope people find our work useful and/or thought-provoking.
Enjoy your journey down the espresso rabbit hole! Even once you feel you’ve got things dialed in, there are a seemingly infinite number of variables to tinker with as you fine tune your shot. A neverending pursuit
Would controlling/adjusting flow rate along with your experiment in tamping pressure make a difference? I say this as the machine I am using (a Decent) is capable of doing this and flow rate seems to affect flavour markedly. The assumption by me is that flavour is correlated with TDS. It seems likely that the same flow rate at the same bar pressure would be more difficult to sustain if tamping pressure was higher in the basket. This makes it difficult for a user to see what lesson is gained from your experiment which isolates brew weight as a significant factor. I just find a take home lesson hard to derive from the experiment although I respect your lengthy attempt.
It’s hard to say the impact of flow rate–and what its relationship with tamping pressure–may be on extraction. We completed this experiment years ago, when the Decent was still in early development. There are many possible confounds that can be introduced in espresso experiments. Perhaps one of the most common ones we typically see is not accounting for/controlling final beverage mass, which significantly dictates extraction. In any experiment, you should design it such that you control as many variables as possible so that the effect(s) from your independent variable(s) can be determined. As with any experiment in any field of study, interpretation of the results should be done with appreciation for the context in which it was collected. The take home for this particular write-up is that tamping pressure, with all other variables being held constant and within the range that we tested, did not significantly affect extraction. That’s kind of all there is to it. A follow-on study systematically assessing the impact of various flow rates at specified tamping pressures on extraction would be interesting. Please post a link here if you do such an experiment!
However … In my personal subjective experience the extraction time depends on the tamping force. I sometimes use coffee up to two weeks after roasting so coffee becomes more stale by the end of the two week period and I compensate for it by applying more and more pressure. I know I should readjust the grinder but I am lazy. It does work.
So if you apply different pressure to the same coffee the brewing time changes and consequently the taste. How it correlates with the amount of solids I do not know. Maybe there is more than solids in the coffee that affects taste.
Improved Puck Density Measurement – Quantitative Cafe
[…] by others, including Socratic Coffee and Tomonori & Smrke, suggests that variations in tamping force have little to no effect on […]
By chance I landed here and what I have noticed is that the pressure you apply does not make a very big difference because when the coffee gets wet, then it begins to expend and the pressure effect is not to notice. Only difference is if you put more coffee in the portafilter and try to fit it in with more pressure. The big difference that I have noticed is with espresso machines where you can change the water pressure: there I saw a big difference with the same coffee with different water pressures. You think that you drink another type or brand of coffee, but it is still the same coffee.
Hi Memo — great observation and comments, thanks for sharing! For this particular effort, we wanted all variables to be the same (e.g., dose of coffee, water pressure) because our goal was to isolate any potential impact of tamping pressure. And, as you have also seemed to notice, within the range we tested, we found no significant difference.
Was this test done with or without preinfusion?
Without. The influence of preinfusion is possibly something to consider, though. Developing an experiment with some standard “preinfusion” definition that is repeatable with the equipment would be challenging (and, as you might expect, any definition of preinfusion you decide on will undoubtedly receive criticism from some folks ;)).
Thanks for the reply. I think it makes sense that without preinfusion, the force of the pressurized water slapping the puck would be far more powerful than any human tamp (9 bar = 130 psi = about 520 lbs of tamp force on a 58mm basket). With a weak 1-bar preinfusion, I imagine there would be a major difference up to a certain weight. I agree that no matter what you’ll get criticism, but it would definitely still be worth it. For science!