This study on extraction was begun by Andy Schecter. His work has gone in a different direction, and he is far less inclined than I to suggest large conclusions on slight evidence; so he has declined being listed as an author. However, the paper owes its existence, its methods, and much of its logic to him. We both started thinking about the subject because several roasters and baristas, including Chris Tacy and Peter Lynaugh were posting on changing how they dosed when using lighter roasted Single Origin coffees. My thanks to Luca Costanzo for telling me about the "French curve" dosing swipers coming into use in Australia and Scandinavia. The move towards SOs in the world wide barista community is leading to a much wider discussion of dosing practices, and I apologize for not citing everyone involved. My thanks to James Hoffman and Alistair Durie, who issued a call for papers via Coffeed.com. Finally, a huge acknowledgment to Ted Lingle, whose classification of coffee flavors by their molecular weight may have provided the best key for tuning the flavor of an espresso shot
The paper presents and discusses data from three related aspects of espresso extraction:
The data sets used in this study are appended at the end.
Some aromas that promise joy from a brewed cup of coffee elicit dread when smelled in a shot of espresso. Why do so many great brewed coffees taste awful as espresso?
The espresso community has long established rules about this. Roughly stated, coffee for espresso has to be low in acidity, not too lightly or darkly roasted, heavy bodied, and contain a significant percentage of dry processed beans. But despite everyone's experience with spectacularly acidic or bitter shots, despite that Pavlovian anticipatory cringe every espresso hound has developed, there is no convincing logical reason for this.
Espresso is more concentrated than regular coffee; so strongly flavored coffees are said to become too strong as espresso shots. But the oils and crema of espresso buffer its extra strength. Moreover, the bad taste of these coffees done as espresso seems less about it being too strong, and more about being unbalanced. The explanation for the unsuitability for espresso of some great brewing coffees may not be in espresso's strength, its extraction of lipids, or its creation of crema. Rather, it may be that espresso brewing extracts flavors differently, and alters their balance in the little cup. This paper tries to get some answers to what this difference is, how it occurs, and how it can be controlled.
summary of results Under-extracted espresso tastes excessively sharp and acidic. Properly extracted espresso has the sweetness to balance the acids and bitters. Over-extracted espresso tastes dull and tarry, or just bitter-sweet.
coffee strength, solubles yield, and taste Everyone knows espresso is stronger than regular coffee; the same tablespoon of coffee brewed into six ounces of regular coffee is used to make one ounce of espresso. But this is not the whole story. Not all the ground coffee goes into the cup or shot. In properly brewed coffee, about 18% to 22% of the ground coffee is dissolved into the water, while the rest is spent grounds; the range in espresso shots is wider, running from 15% to 25%. This proportion of ground coffee that ends up in the cup is called the solubles yield; a brew with too low solubles yield is called under-extracted, while one with too high solubles yield is called over-extracted.
It is important to understand that solubles yield and brew strength are two separate things. For instance, imagine that the six ounce cup of brewed coffee was extracted at 24%. Its strength would be 24% of the coffee measure divided by 6 ounces or roughly 4% per ounce. If the same measure in a shot of espresso was extracted at 16%, all of it would still be in one ounce of water, so the under-extracted espresso would still have four times the strength of the over-extracted brewed cup(1).
If we imagine tasting the same amount of extraction in less and less water, we would notice the same aromas and tastes getting stronger and stronger. In an analogy, it's the volume getting louder, but the music staying the same. How about if we taste more and more extracted ground coffee in the same strength of brew? Roasted coffee has over a thousand different flavor chemicals. Some of these dissolve quickly, some more slowly. As the grounds become more extracted, the slower dissolving flavors become more prominent in the brew. So, in our analogy, it would be the volume staying the same, but the music gradually changing.
This paper is about the solubles yield and how it a affects the taste; it is not about brew strength. The brew strength of a standard espresso shot has about 120 parts per thousand coffee solids. The shots I made for this paper ranged from about 100 to 160 parts per one thousand, with weaker and stronger shots occurring at all solubles yield levels. In the data analysis, the effect, if any, of the brew strength, is removed from the results.
how to taste for proper extraction With a thousand flavor compounds to play with, no two coffees are, or ever will be, the same. So how can there be any way of knowing what extraction is best for what coffee? Fortunately, these flavors can be grouped into a few large families, so that all the members of a given flavor family extract in similar ways. This work was done by Ted Lingle, who grouped the flavors by molecular weight, with the light weight ones dissolving quickly, and the heavy weight ones dissolving slowly:
There are two light weight, fast dissolving, families that are fully present even in under-extracted coffee.
There are also two heavy weight, slow dissolving families which require high solubles yields to reach their full strength.
This flavor classification(2) provides a road map to a balanced coffee extraction, either for brewing or in espresso. While describing the taste of coffee both accurately and in detail is an art; it is fairly easy to sort the tastes and smells into these four broad groups. It is even easier to tell if the coffee is under-extracted, properly extracted or over-extracted:
I had previously worked out a way to measure and control the solubles yield in espresso by varying the dose and grind (see next section). I was aiming for extractions from 16% to 24% solubles yield. To get this, I used an LM triple basket, and did shots at 12 grams, 14.5 grams, 17 grams, and 19.5 grams, grinding each dose so get normal double shots in around 25 to 35 seconds. I did six shots at each dose level, and measured the extraction level.
Four of the shots at each dose were done with the City (Northern Italian) roast, two at a Vienna (Southern Italian) roast. The coffee was a 50/50 mix of Idido, a DP Yrgacheffe with floral, citrus, chocolate, and green tea flavors, and Cenaproc, a WP Bolivian Bourbon with an apple acidity and very sweet marzipan caramels. I rated the aroma, mouthfeel, and crema appearance of each shot on a zero to ten scale. I also rated the balance of acidic to bitter flavors, with zero as most acidic, ten most bitter, and five as neutral. Finally, I rated the overall sweetness from zero to ten.
There were two flaws to this test; so these results need to be confirmed.
The crema, mouthfeel and aroma had no relation to the solubles yield(3). It is possible to get excellent and alas lousy shots in these aspects at all yields
As the extraction model predicts, higher yield shots tended to be sweeter than lower yield ones for all roasts(4); but lighter roasts responded to yield differences more dramatically than dark ones. This is logical, since dark caramels are not as sweet as light ones. At very high yields, the light roast shots became distinctly caramel flavored, indicating over-extraction. Extractions around 23% for the light roast, and around 20% for the dark one tasted the best balanced
Higher yield shots tended to a more bitter balance, and lower yield ones to a more acidic balance(5). This was equally true for light and dark roasts, although the darker roast was of course more bitter over the entire series. This effect would not have occurred with a light roasted, low acid coffee dominated by light maillard flavors, such as a Monsooned Malabar. In such a coffee, the taste would have been bitter at all yields, with increasing sweetness at higher yields
While this taste test is hardly conclusive, it does support the relationship between solubles yield and taste implicit in Ted Lingle's work on the flavor chemistry of coffee. It should encourage people who enjoy espresso to experiment with varying the solubles yield in their shots, and to confirm that the shot's taste can be tuned by doing this.
So how does one control the solubles yield?
summary of results For a given basket, grinder, and machine, higher doses lead to lower extractions, lower doses to higher extractions.
measuring solubles yield As usual, Andy Schecter had asked the right question and done the right thing. He asked "How do you know a ristretto is really stronger than a normale? Maybe the ristretto has less of the puck in it." Then he started weighing shots and pucks.
I was obsessing about the extraction process when he emailed me about his project, and I realized he had figured out a way to get control over the solubles yield in extraction. The yield is the percentage reduction in puck weight as its constituents are dissolved into the espresso. If one weighs the puck before and after the shot, one can figure it out. If one does it over and over again, for different baskets, doses, coffees, shot times, shot weights, and puts the data through the statistical meat-grinder; one can detect the patterns and learn to control the extraction. I was off and running.
The puck weighing has some additional requirements. The puck is wet after the shot, and needs to be oven dried before weighing. Fresh ground coffee holds some water, so a grind sample also needs to be oven dried and the initial puck weights have to be adjusted accordingly. Finally, some grounds from the puck cling to the shower screen and group gasket after the shot; these need to be recovered with careful brushing.
There are two sources of error in the yield results. Any ground coffee not recovered for the weighing would reduce the puck's weight and register as increased yield. The puck's moisture contains extracted coffee which is added to back to the puck's weight while baking, thus registering as decreased yield. Marino Petracco, in the extraction chapter in Illy, cites 25% yields for Italian dosing standards. My measurements were about 2% to 3% less and may be an artifact of this soggy puck error. However, since these sources of error are the same for every observation, the relationship between yield and its predictors is unaffected.
controlling solubles yield My observations ranged from 15% to 25% of puck weight extracted. There were two outliers, one at 12% and one at 26%, both coming from extractions well outside normal espresso time and volume limits. The next task was to find out which of the controllable shot parameters can be used to set the yield.
Grind setting was not consistently related to yield except when using the same basket, coffee, and roast level. Different coffees, roasts and baskets require different grind ranges.
The major predictor of yield is the weight of the puck divided by the hole area at the bottom of the basket (P/A). I measured the diameter of the hole-punched area at the bottom of the basket and squared it. This figure is not the actual hole area; but is directly proportional to it and easy to determine. The P/A measure can be simply understood as the depth of the puck in the basket; a fat or deep puck extracts less solubles than a thin or shallow puck. When the puck is conical rather than cylindrical, so that the top surface is larger than the surface against the filter holes; the P/A measure can be understood as the average length the water has to travel through the ground coffee. It is operationally identical to the depth of the puck in a perfectly cylindrical basket.
This straight line relation explained about 75% of the variations in yield, and the predictions had a standard error of 1.8% yield level. Moreover, the formula using P/A remained the same for all baskets, coffees, and roast levels for the same machine. In the graphic below, which applies to the Elektra Semiautomatica group, an LM double basket dosed at the Italian standard of 12 to 14 grams would yield about 22% to 20%, and only 17% to 18% when dosed at the US standard of 17 to 19 grams.
Andy's observations with his highly modified Silvia also showed the P/A relation with the same predictive power. However, the slope of his observations is different from mine. This graph shows only the readings I took on the Elektra
Changes in shot time and weight also affected the yield, but the effect, when shots were cut as they blonded, was surprisingly small(6). Adding these variables to the predictor equation only improves the model's overall accuracy by about 3%, and only reduces the standard error in the yield prediction from 1.71% to 1.65%. Given the added complexity of the formula, it has no everyday use.
The first reaction I got to this work is that low dose shots probably taste weak when compared to high dose shots. This is true when a poorly trained barista pulls a shot with roughly the same volume for both doses. If one attends to the color of the stream when stopping the shot, the solids concentration in the cup remains fairly constant regardless of dose. In this experiment, in the 24 shots I made for the taste test, there was no relation between dose and the amount of solubles in the cup. In the graphic below, shot concentration is measured as the weight reduction of the puck (solids getting into the cup) divided by the weight of the shot. The unit is parts per thousand, and the dashed line shows the 120 parts per thousand standard.
In practice, on any given machine, one can use a simple straight line P/A formula to calculate the yield for any basket and coffee. However, since it is different from machine to machine, and maybe grinder to grinder, all baristas will have to work it out for their own setups. Fortunately, it is not required to know the exact figures. One only needs to know that going to a lower dose and finer grind in the same basket increases the extraction. One can taste the shots, analyze their flavors using the taste model in the last section, and make dosing changes in the right direction until one has found the correct dose for every coffee(7).
Why does it work this way? Why does only the dose and basket shape matter a lot, while the shot time and volume don't matter much? To answer these questions, one needs to learn more about how the coffee extracts during the course of an espresso shot.
summary of results The yield does not depend a lot on shot time or weight, because the puck is extracted almost as far as it will go within the first 20 seconds of the shot. I'm not sure why the total amount extracted is nearly done by twenty seconds, or why it varies by dose. That doesn't stop me from speculating about it at the end of this section
the taste of the early, middle and late part of an espresso shot There's an old alt.coffee exercise: brew an espresso shot into three cups; the first 10 seconds into the first cup, the next 10 seconds into the second cup, the rest into the third cup. Then taste(8).
When we discussed this, we assumed that the puck was extracting evenly throughout the shot; so that the first cup represented the 0% to 7% solubles fraction, the next cup the 8% to 14% solubles fraction, and the third cup the 15% to 21% fraction. However, the last section shows that the degree of extraction is relatively immune to shot time and weight. This leads to a, in hindsight, much simpler interpretation of the three cups data. The first cup represents the early extraction, the second cup the late extraction, and the third cup a diluting of the shot. If the puck loses almost all its solubles into the first 20 seconds and 50% weight of the shot; then the near immunity of the yield data to shot timing is explained. All the shots in the extraction data set were within normal espresso parameters. So the only thing that varied is by how much the full extraction, present in the cup after the first 20 seconds, was diluted by the rest of the shot.
Is there any way to confirm this supposition?
intra-shot solubles yields The obvious solution is to check the solubles yield of the puck at various stages over the course of a single shot. This can't be done with just one shot, but it can be done with six shots stopped short at staggered intervals which use the same coffee, dose and grind. I did two of these sets, stopping the first shot in each set at the first drop, and the subsequent ones at 6 second intervals after that, out to 30 seconds (actually 35, since the dwell time is around 5 seconds). I used a high dose, because I was also disecting the puck (see below), and the final yield is, as predicted, on the low side. The graph shows the averaged readings of the two sets(9).
This confirms that the bulk of the solids extraction takes place in the first 2/3rd of the shot, and begins to explain why shot time and weight play such a minor role. What it doesn't explain is:
getting inside the puck while it is brewing I haven't gotten very far with these questions; and I may be looking in the wrong direction. But, I think the answer lies in what happens inside the puck during the shot. This is pretty much a "black box," but with some theory and some tricks, one can begin to crack into it.
The theory is based on the idea of percolation. This has nothing to do with those accursed coffee percolators from the 50s. Rather, it looks at the brewing process as water flowing through a column of ground coffee. Since the puck is, sort of, a column of ground coffee, and since the solubles yield depends on how high this "column" is, it seems an appropriate way to think about the problem.
Discussions of percolation(10) can get confusing very fast. The ground stuff is "coffee," the liquid coming out of the bottom is "coffee," the stuff going from the grinds to the liquid is "coffee." One needs to come up with a new vocabulary to prevent total confusion. Here goes:
The mental model percolation is real simple. The liquid goes into the top, picks up solubles from the grinds at the top, and becomes saturated. Once it's saturated, it can't pick up any more solubles, so the solubles stay inside the grinds at the bottom until the top grinds are exhausted, and the liquid reaching the bottom is less saturated. In other words, the percolation column brews from the top down.
The problem is that this mental model doesn't fit our facts. In this model, it doesn't matter how high the column is, or how coarse the grinds. It brews from the top down. Send through enough water, and it will extract as far as you desire. But the intra-shot graph shows that the extraction basically levels off after 20 seconds. The dose by extraction data shows that the amount of extraction in normal shots varies almost entirely by puck height, and very little by time or volume.
Time to gather some data and make the model more complicated. The data is a series of shots stopped at the 6 second intervals; with each puck divided into three horizontal slices; and each slice's brew strength measured. The more brew strength, the less coffee has been extracted. This provides a picture of how the top, middle and bottom of the puck extract over the course of a shot.
The puck sections were oven dried, and brewed at exactly 4 grams powder to 80 grams of water. These brews were compared to the fresh coffee, also oven dried, brewed at 4 grams (100%), 2 grams (50%), and 1 gram (25%) per 80 grams water. The comparison was by TDS meter. I did two series of measures.
The three pairs of horizontal lines show the TDS of the fresh coffee at 100%, 50% and 25% brew strength. They provide a rough measure of the extraction levels of the puck sections, with the 25% line indicating a full extraction. The top of the puck is shown by the blue lines, the middle by the green lines, the bottom by the red lines.
The measure at time zero is at the first drop. It shows the puck state when it gets completely soaked. And it leads to out first real world revision of the simple percolation model: grinds absorb liquid. The liquid absorbed at the top is water, while the liquid at the bottom has all the solubles it picked up at the top. So, when the percolation column becomes completely soaked, but before any liquid comes out, solubles have been transferred from the grinds at the top to the grinds below. Below a certain depth, the grinds actually are charged with solubles, rather than losing them (the same trick is used to recharge decaf coffee after the caffeine has been removed)
The next revision comes from looking at the shape of the lines. They are roughly straight, rather than the asymptotic curves one expects from the simple model (the intra-shot extraction graph in the previous section is an asymptotic curve, it reaches a limit). The water is not marching down the puck column in a straight line. The puck has become soaked, and has turned into a slurry. The fresh water enters this slurry and mixes with it. The mixed water is pushed out of the bottom. The proper mental model is of a big pail filled with liquid and grinds with a filtered hole in the bottom where liquid but not grinds flows out. The water in the pail is both brewing and being diluted by the in flowing water. It's the mixed up result that emerges from the bottom.
Do these two revisions to the percolation model explain the extraction results?
A bit; but not like a slam dunk. In the soaking phase, the thinner the puck, the less overloaded with solubles the bottom layer becomes. Since the bottom layers can't extract their own solubles until they've got rid of the excess ones from the top, thinner pucks help along the extraction. In the progressive dilution phase, the bottom of a thinner puck will get more fresh water and extract further. These two factors together may explain the results.
There's a lot more to the story that I don't know or can't document. Thinner pucks require finer grinds, so the extraction is probably accelerated by that. The fines, and finer particles in general, migrate toward the bottom of the puck, so when the extraction at the bottom gets going, it probably proceeds faster than at the top. Top down extraction, soaking and dilution phases, thinner and thicker pucks, finer and coarser grinds: there's a lot happening during an espresso shot. These intra-shot measures are a first peek, but there's a lot more looking required.
None of these complications should be allowed to obscure the basic point: the taste of espresso varies by the level of extraction, and that level can be controlled.
And this result points to the irony of unintended consequences.
In Italy, espresso is a mass consumption item, mostly made from coffees of the same low quality as is found in supermarkets everywhere. Since the aromas of such coffees are not all that great, staling is of little consequence, while keeping doses precise and yields high is of great consequence, since one needs to extract every iota of caramel to make the shots palatable. So the ground coffee sits in dosers going stale, but is precisely dosed, 6.5 grams into single baskets, 13 into doubles.
In the non-Mediterranean world, espresso is specialty coffee(11). Cafe owners rightly noticed that the ground coffee was going stale in the dosers, and went to alternative dosing methods. What they didn't notice is that dosing by leveling the freshly ground coffee to the basket's rim, a dose far higher than the Italian norm, gets solubles yields of 16% to 20%, rather than the 20% to 24% one gets with a properly adjusted doser. "Specialty Espresso" was almost always under-extracted
At these low extraction levels, high grade coffees become a bane rather than a boon, producing jarringly acidic or sharp shots. So, in the specialty coffee world, there is a feverish search on for ultra-sugary high grown coffees that are still sweet when roasted light and under-extracted. When such coffees are not available, one gets the ubiquitous medium-dark roasted blends that are a far cry from the quality of the specialty coffees sold for regular brewing.
And all this because of an unintended consequence of using fresh coffee. I think it's high time for baristas to relearn their dosing.
As a final note: since posting early drafts of this paper, I have found out that this is changing already. In Scandinavia and Australia, many top competing baristas are replacing their fingers with curved swipers that scoop out ground coffee below the rim level of the basket. By having a french-curve like set of these, they can efficiently vary the dose in a workplace or competition context. I'm sure these "3rd wave" dosing tools will become much more prevalent and developed as word on working the solubles yields gets out.
(1) For a more detailed explanation, consult Ted Lingle, The Basics of Brewing Coffee, 1996, SCAA.
(2) The classification used here differs from the one in Ted Lingle, The Coffee Cupper's Handbook, 2001, SCAA. He divides flavors into three main groups, enzymatic, sugars browning, and dry distillates. However, the herby and nutty sub groups contain flavors that derive mostly from Maillard reactions between sugars and amino acids. These are the flavors that typify toasted grains, malts, smoked meats and barbecued foods. I believe they deserve their own overall classification for three reasons: they are associated with sharp or bright-bitter tastes, rather than the sour or sweet tastes of the neighboring groups; they have similarly fast solution rates; and they are mostly produced in the part of the roast running from 300F to the first crack, and hence can be roast-profiled as a group.
(3) Aroma slightly improved at high extractions, but the effect is only marginally significant (0.95):
Mouthfeel and crema are, as expected, well related to the concentration of the shot. Here are the relevant partial regressions:
(4) The residual correlation of sweetness to yield, after removing all other factors, was 56%, with a t-value of 3.12.
(5) The residual correlation of acid/bitter to yield, after removing all other factors, was 73% with a t-value of 4.82.
(6) The t-value for the simple linear regression of filter.area/dose to yield is 7.89; if one uses the best predictor: filter.area/dose*log(shot.time*shot.weight) the t-value rises to 8.69. The graph is shown with the predictor inverted to the more sensible dose/filter.area, so the regression line is transformed to a hyperbolic curve. With t-values this huge, the magnitude of the change in the correlation coefficients from 75.2% to 78.3% has a high degree of certainty, and shows that while shot time and weight do expalin some of the yield changes, they definitely don't explain a great deal.
The small effect of shot time and weight/volume is probably an artefact of cutting shots when the flow blondes. This cuts fast flowing shots short, and lengthens the slow flowing ones. This practice counteracts the physically mandated higher extraction rates of fast flowing shots. However, since cutting a shot as it blondes is proper barista technique, the result applies to actual shot making.
(7) Dosing changes can also affect the shot because higher doses can come into contact with the shower screen, while lower doses do not. This depends also on the height of the basket. M. Petracco warns against puck contact with the shower screen in chapter 7 of Illy and Vianni, eds, Espresso Coffee: the Science of Quality, 2005, Elsevier.
On my Elektra, the shower screen is like a third rail, so all the data in this paper are from shots with head space. In other groups, shots seem to survive compression by the shower screen without harm.
(9) The 12 second reading only contains one observation, since I messed up the other one.
(10) Chapter 7 in Illy, as cited in note 7 is a discussion of espresso percolation. It focuses mainly on how the puck dynamically affects the flow, and helped me appreciate the important role of the grinds absorbing liquid. However, it does not go into the timing of solids extraction. Discussions of percolation for instant coffee, pp 127-128, Clarke and Vitzthum eds, Coffee: Recent Developments, 2001, Blackwell Science, are frightening; but they emphasize the role of how the initial percolation column gets wet (top down versus before starting the percolation), and how hard it is to get a fully extracted or in the instant coffee case an insanely over-extracted, output.
f.loss puck.b puck.a p.loss sh.wt sh.time aroma ac.bi dr.sw body crema roast 1 0.962 11.8 8.7 0.234 22.7 28.0 8.50 4.00 6.50 7.00 8.50 1 2 0.962 14.8 11.5 0.192 23.2 35.5 7.50 3.75 6.75 7.25 8.50 1 3 0.962 17.9 14.2 0.175 21.9 37.5 8.00 3.25 5.25 7.00 8.75 1 4 0.962 20.7 16.3 0.181 30.8 22.0 8.00 3.00 3.75 7.25 7.50 1 5 0.976 12.0 8.8 0.249 18.4 28.2 8.00 4.50 7.00 7.50 8.50 1 6 0.976 14.3 11.3 0.190 21.0 35.0 8.50 3.75 5.75 7.00 8.50 1 7 0.976 16.9 13.2 0.200 26.7 28.4 8.00 3.75 4.75 7.50 8.00 1 8 0.976 19.6 15.9 0.169 30.5 23.2 8.00 4.00 5.50 7.00 7.50 1 9 0.975 12.0 9.0 0.231 26.2 29.4 8.50 4.50 7.00 7.00 8.50 1 10 0.975 14.1 10.9 0.207 26.4 31.2 8.00 4.25 6.50 7.50 8.50 1 11 0.975 16.6 12.9 0.203 31.5 41.2 7.50 3.75 5.75 7.50 8.00 1 12 0.975 19.5 15.6 0.179 30.5 36.4 7.50 3.00 5.00 8.00 7.50 1 13 0.975 12.0 8.9 0.239 25.3 30.0 8.50 3.75 6.50 7.00 8.50 1 14 0.975 14.1 11.1 0.193 23.3 30.0 9.00 3.50 7.00 7.00 8.50 1 15 0.975 16.6 12.8 0.209 47.7 30.0 8.50 3.75 5.00 6.50 7.00 1 17 0.981 12.0 8.9 0.244 22.2 58.0 7.50 6.50 7.00 7.00 8.50 3 18 0.981 14.4 11.3 0.200 25.0 53.0 7.00 6.00 6.50 7.50 8.50 3 19 0.981 17.0 13.6 0.185 24.0 42.0 6.00 5.75 5.75 6.50 8.00 3 20 0.981 19.5 15.4 0.195 38.4 30.0 6.00 5.75 7.00 6.00 7.00 3 21 0.981 12.0 8.6 0.252 25.8 36.0 8.00 7.00 6.50 7.50 8.25 3 22 0.981 14.5 12.0 0.194 19.7 51.0 6.50 4.50 5.50 8.00 8.50 3 23 0.981 17.0 13.7 0.204 33.0 46.0 8.50 5.25 5.25 7.00 8.00 3 24 0.981 19.5 15.5 0.173 22.9 32.0 6.75 5.00 5.75 7.50 7.50 3
dose yield sh.wt sh.time filt.d roast 1 10.5 0.174 9.1 45.0 2.9 2 2 10.3 0.168 14.1 59.0 2.9 2 3 8.4 0.186 22.9 25.0 2.9 2 4 8.0 0.171 12.0 28.0 2.9 2 5 8.3 0.189 22.0 36.0 2.9 2 6 17.6 0.194 17.8 39.0 4.3 2 7 18.5 0.189 18.0 32.0 4.3 2 8 19.5 0.205 20.1 55.0 4.3 2 9 14.8 0.214 28.4 43.0 4.3 2 10 20.4 0.190 17.5 52.0 4.3 2 11 17.0 0.196 17.8 36.0 4.3 2 12 9.9 0.155 8.1 24.0 2.9 2 13 9.8 0.157 11.5 35.0 2.9 2 14 21.4 0.180 18.9 26.0 4.9 2 15 18.2 0.181 16.2 43.0 4.3 2 16 13.7 0.203 31.3 26.0 4.3 2 17 18.3 0.197 23.3 32.0 4.3 2 18 23.5 0.175 21.3 65.0 4.9 2 19 17.2 0.187 15.7 60.0 4.3 2 20 16.7 0.206 26.4 31.0 4.3 2 21 10.2 0.180 16.7 26.0 2.9 2 22 17.8 0.160 15.7 38.0 4.3 2 23 10.9 0.132 8.7 36.0 2.9 2 24 10.0 0.156 15.1 26.0 2.9 2 25 18.2 0.178 20.7 47.0 4.3 2 26 11.8 0.234 22.7 28.0 4.9 1 27 14.8 0.192 23.2 35.5 4.9 1 28 17.9 0.175 21.9 37.5 4.9 1 29 20.7 0.181 30.8 22.0 4.9 1 30 12.0 0.249 18.4 28.2 4.9 1 31 14.3 0.190 21.0 35.0 4.9 1 32 16.9 0.200 26.7 28.4 4.9 1 33 19.6 0.169 30.5 23.2 4.9 1 34 12.0 0.231 26.2 29.4 4.9 1 35 14.1 0.207 26.4 31.2 4.9 1 36 16.6 0.203 31.5 41.2 4.9 1 37 19.5 0.179 30.5 36.4 4.9 1 38 12.0 0.239 25.3 30.0 4.9 1 39 14.1 0.193 23.3 30.0 4.9 1 40 19.5 0.127 18.0 30.0 4.9 1 41 12.0 0.244 22.2 58.0 4.9 3 42 14.4 0.200 25.0 53.0 4.9 3 43 17.0 0.185 24.0 42.0 4.9 3 44 19.5 0.195 38.4 30.0 4.9 3 45 12.0 0.252 25.8 36.0 4.9 3 46 14.5 0.194 19.7 51.0 4.9 3 47 17.0 0.204 33.0 46.0 4.9 3 48 19.5 0.173 22.9 32.0 4.9 3
1st Set time TDS.top TDS.mid TDS.bot STR.top STR.mid STR.bot [1,] 0 663 1112 1424 65 90 120 [2,] 6 650 798 1250 60 60 110 [3,] 12 440 815 1235 40 50 110 [4,] 18 478 571 1041 45 45 75 [5,] 24 409 512 1004 40 45 75 [6,] 30 321 357 690 25 30 55 2nd Set time TDS.top TDS.mid TDS.bot STR.top STR.mid STR.bot [1,] 0.0 854 1248 1676 75 100 120 [2,] 6.0 542 971 1595 40 70 100 [3,] 13.5 578 781 1180 45 55 85 [4,] 21.0 373 643 1113 25 50 85 [5,] 25.5 406 674 1058 35 50 85 [6,] 30.0 292 487 709 20 35 60