SLIDE 1
Table of Contents Section 1.01 Questioning the history of technical - - PDF document
Table of Contents Section 1.01 Questioning the history of technical - - PDF document
Table of Contents Section 1.01 Questioning the history of technical standards........................................................... 2 (a) Rational for the
SLIDE 2
SLIDE 3
3 moments earlier without tamping at all. Speculation in response to his effort ran the range of commentary, from simply crediting his experience to whether or not he had secretly tamped the coffee out of view in some stealth-like manner before mounting the filter. It was the former. The history of these assumptions is hard to track down. When presenting machines to Starbucks in early 1992, their technical department took substantial space, was full of machines from virtually every manufacturer and staffed by two ex-baristas with no training in engineering. Even though most of us like to think we have risen above the mass market appeal of Starbucks, the corporation that had developed the gold standard for promoting espresso as we know it in the United States was also responsible for driving much of what we assume to be true of espresso machine standards. This from a company that, at the time, employed no engineers and had no understanding of the process of pre-infusion, temperature profiles or the importance of metal composition in the constructions of espresso machines.
Section 1.02 Comparison between the United States and Italy
(a) The United States lacks the institutional background
I don’t expect Italian manufacturers to produce great American coffee brewers. They have tried, and quite frankly, they aren’t very good. Bunn, Curtis and Grindmaster employ engineers who are perfecting the process of brewing American coffee based upon a legacy of decades in the industry. Each new technical innovation: profile brewing, interfacing grinder to brewer and solid state controls rest on the shoulders of the engineers and coffee professionals who came before them. The best assets we have here in the U.S. are actually the growing ranks of consumer members who have brought their backgrounds in other fields to their love of coffee. Greg Scace brought his experience in scientific measurement processes to espresso machines as a means of understanding the process of extraction and we have experts in process control and PID programming regularly providing insight into espresso theory. It is often these consumer members that are unraveling the myths and secrets of great espresso; what they lack in professional espresso experience they make up for in a rigorous search for the answers.
Luigi Lupi, from Elektra. Marzocco has, by far, the best badge holders
SLIDE 4
4
(b) The Italian approach to technology (i) Institutional memory
With few exceptions, we in America lack this institutional history. Our efforts rely upon trying to understand the technology with which we are presented. We have consultants, authors and coffee bar owners trying to unravel the technology developed by engineers in Italy who have worked for decades to perfect the art of espresso. Italians first patented espresso machine technology in 1906 and have built an industry to refine and perfect the process. The same historical perspective of how best to brew American coffee here is represented in Italy. The engineers I have worked with more often than not assess a machine by looking at the espresso, observing whether it’s too hot or cold, or whether the pressure is set improperly by the characteristics of the shot. Years of experience have replaced the need for testing. I have seen this lack of enthusiasm for the particulars of espresso machines on the part of Italian representatives interpreted by American observers as a lack of concern for the technology; that the Italian attitude about refining and perfecting the process of extraction has become lax and
- unconcerned. Noting could be further from the truth.
(ii) Scientific approach and investment
Espresso machine manufacturers spend hundreds of thousand of dollars each year developing new products and refining old processes. Secretive offices manned by engineers with actual degrees spend countless hours over the course of many years working out the details of a particular group design or heat exchanger configuration. One of the many engineers I worked with was previously employed by Magnet Marelli designing the electronics for Ducati Motorcycles. Another engineer was working on a new material composition for an
- -ring to increase the longevity of an automated group design.
Many of the efforts we take for granted, not because these engineers aren’t concerned about the quality of the espresso their machines produce, but because it is masked by everything else necessary to produce a reliable product that interfaces well to the end user.
(iii) Alternate concerns
Espresso machine manufactures often have competing interests to the simple mechanical process of making espresso. They also have to look good, as they are often the primary feature of an Italian espresso bar. They must also be reliable and viable to produce. From gold plated, eagle perched works of art to simplified piston driven machines, most manufacturers offer a range of machines suitable for a variety of markets. What we see here in the United States is often a small subset of the production from a typical manufacturer.
SLIDE 5
5
Section 1.03 Exploring technical standards
(a) Temperature (i) Origins of temperature stability theory 1) Easy to understand, not necessarily gospel
The notion of trying to achieve temperature stability in an espresso machine is an easy
- ne to grasp; in part, because it just makes so much sense. We know that there is a relationship
between the temperature of water and the quality of the extraction, it follows that the more precisely we can control the temperature of the water, the better we can more consistently reproduce the process. This would make sense if temperature was the only factor involved with producing a quality shot of espresso, but it isn’t. Different machine dictate separate temperature profiles for identical coffees; Café Mauro in Italy specifies a separate temperature setting for each of about a dozen machines. All this for the same blend of coffee and it varies by as much as 3 degrees Celsius—a significant amount. If temperature alone was the determining factor for reproduction to a roaster’s standard, it should be the same for every machine. It doesn’t. The manner in which a machine handles the process of extraction differs enough that different temperature profiles are required by each machine. The static temperature of the head, dimensions of the portafilter, filter bed thickness, temperature profile curve and Gicleur orifice can all affect the extraction process and do so simultaneously. To choose one factor among them all as the single most important, while easy to analyze, is also simplistic.
2) Italy’s quest for balance in the relationship between pressure and temperature
Whether by design or by accidental observation—it has been too long to know for sure—Italian engineers have worked to understand the interaction between temperature and pressure and its effect on coffee. The earliest machines were little more than automated stove top espresso makers, using the steam pressure to force water through finely ground coffee. Early attempts to increase the pressure resulted in higher temperatures that burned the coffee. Only when a large spring was provided to press the preheated water through the coffee was a reasonable solution found. The added benefit of this design was that low pressure pre-infusion effect with elevated water temperatures during the initial stages of the extraction.
SLIDE 6
6 The term often used to describe this elevated temperature at the beginning of the process is ‘crispy water,’ which probably looses something in translation. I have not found any reference to the effect in literature, I suspect, because it is often taken for granted as an accepted feature of many machines. The degree, duration and profile vary with each machine. In as much as a car manufacturer is unlikely to reveal exactly what method they use to port an intake manifold to produce a particular level of torque, espresso machine manufacturers are unlikely to publicize their own processes. Little is patented in Italy and most are very proud of their own take on the process—leaving it up to roasters to decide and specify a particular profile.
(ii) Temperature profiling 1) American Models
The following represents an ideal scenario for extraction according to WBC standards, a relatively flat and predictable temperature profile using a single boiler dedicated to each group with a preheating system to insure stability.ii
SLIDE 7
7 The graph below is a close up of the shot demonstrating the stabile temperature achieved with this system. While this is quite an achievement, the history for validating the desirability of temperature stability is anecdotal. The premise is untested in side by side qualitative testing with competing profiles. This may produce a great shot of espresso, and it may be able to be adapted to a more customized method of profiling by modifying the input water temperatures. There has to be an incentive to do it; which isn’t likely yet.
2) Italian profiles, Typical
The graph below was generated by Ken Fox, whose experimentation with his Cimbali Junior has become legend among espresso machine modders. Though this machine has been
- utfitted with a PID controller, the temperature profile is unchanged from the original design
and, according to Ken, is not significantly different in terms of repeatability from the standard pressure control used in the machine. The temperature hump at the beginning of each shot is typical of an Italian temperature profile design. The amplitude and duration of the profile is dependant upon the system design coupled with the method of pre-infusion used. Larger volume pre-infusion systems can accommodate larger and longer initial profiles. The argument among the Italian engineers is that the higher temperature elevates the puck temperature more rapidly and opens the coffee— though it’s not clear exactly how that translates. Because the initial extraction pressures are low, the water penetrates the coffee without extracting much of the liquor and soluble
SLIDE 8
8
- compounds. The process Ken used was designed to accommodate what a typical home user
would encounter, but is typical for a machine that would be used regularly in a commercial
- environment. iii
The following demonstrates the subtle changes that can occur in the temperature profile when usage of the machine changes. The lag times between shots were greater—about 10 minutes—with similar results; elevated initial infusion temperatures with quite stable temperature profiles.
SLIDE 9
9 This kind of profiling requires that each group have an independent environment within which to work. On typical heat exchanger based machines, that means an isolated system with its own profiling capability. European manufacturers (as well as one in the US) use this model. Machines are never in constant use—though they may appear to be—as the time it takes to remove and reload the next shot approaches the time needed to recover temperature and begin a new thermo-compensation cycle. This period of use and rest between shots forms the recovery cycle for the group. The flow rate of the heat exchanger in thermo siphon systems (more often now adjustable by the end user/roaster) and the length of the cold water injection tubes determine the recovery rates and repeatability of the profile. Cimbali uses additional tubes through which cold water passes in the head to tune temperature profiles. Most bars in the US never encounter situations where usage over runs these cycles of recovery, considering bars in Italy routinely pull several thousand shots a day. The systems are designed in such a way as to minimize temperature fluctuations at the head, even under typical demands for steamed drinks in the US. How these machines maintain static temperatures in a dynamic environment is evolving, but the heat exchanger designs act as a buffer between the temperature fluctuations in the boiler steam jacket. Temperature stability relies on a number of factors. The design of the system itself can have a built in mechanism to adjust the flow through the brewing heads based upon the temperature differential between the boiler and the brewing group. Placement of the exchanger in the steam jacket can further stabilize the temperature as temperature fluctuations in the upper portion of the boiler are not as dramatic as they are adjacent to the heating element. The important thing to consider is that the water in the heat exchanger is not heated by the large surface temperatures fluctuations of the heating element, but by the much more stable wall temperatures of the heat exchanger itself.
(iii) Methods of temperature control 1) Pressure stats
Pressure controls have been the standard mechanism for temperature regulation on most modern espresso machines until quite recently. Even with the advent of electronic control designs, they offer some significant advantages. The relationship between temperature and pressure is very accurate, linear and predictable. The rapid response of the controls is a result of the average boiler temperature sampled by the expansion and contraction of steam
- pressure. The reactions to fluctuations are virtually instantaneous, such as when steam is
removed from the boiler. The disadvantage is that is that there is no proportional control or metering of power to the element to maintain a more constant temperature. Pressure stats also cycle on and off over a span referred to as a dead band. The range of the band is typically 0.14 bar, or about 3º Fahrenheit on a new spec pressure stat. The narrower the dead band, the tighter the fluctuation—however rapid cycling of the element at full power could shorten the life of both the switch and the element. New, narrower band pressure stats with solid state controls improve the response curves without damaging the relays controlling the elements.
SLIDE 10
10
2) Electronic controls/PID
New electronic controls are being integrated into machines as either part of the electronics or as separate devices. Often referred to as PID controllers (P -Proportional, I - Integral, D – Derivative), they are able to modulate the heating element, providing just enough power to maintain very precise temperatures. They can program themselves with a built in learning curve adjusting to the demands needed for each heating element and boiler size. They can also be programmed to override those settings to be more or less responsive to changes in the boiler environment. With PID controllers, it is possible to maintain boiler temperatures to within a few tenths of a degree. So long as you don’t steam any milk, draw any water or make too many shots. While the program necessary to maintain temperature is very precise, it is not the same program necessary to respond to rapid changes in boiler temperature that occur when steaming milk or when the boiler refills with fresh water. By the time that the PID controller has sensed a dramatic change in temperature enough to call for full power to the heating element, it may have already dropped below the temperature at which a pressure stat would have come in at full power, as much as 6º Fahrenheit. The program necessary to accurately recover from such a drop is also different from that required to maintain temperature and the controller may over- shoot the set point slightly, or be too cautious (depending upon the program) and stabilize too slowly. Sensors are also a problem for temperature controllers. They have their own mass and steam is not the best conductor of heat energy. The sensor itself may lag behind the actual boiler temperature before sending a signal to the PID. These considerations make PID controllers more ideally suited to dedicated boiler designs, where the temperature fluctuations from making espresso are more predictable and less dramatic. The inability to respond to demands, even in these environments, has still been an issue. The most recent dedicated boiler machines have resorted to preheating the water that feeds the brewing boiler so that the PID can work towards doing its designed task of maintaining temperatures as opposed responding to external changes.
3) 3 PID based pressure controls, fuzzy logic and predictive technology.
PID controllers don’t have to just control temperature, however. Than can be used to pressure as well. Because the response to pressure fluctuations is quite fast, the latency of the temperature sensor can be eliminated as part of the control process. While the PID controller would have to be tuned to better respond to the rapid changes in pressure, a controller set up this way would actually be more responsive to steam demands yet provide the stability of boiler temperature Americans are looking for. To truly regulate temperature, should that be the ultimate goal, fuzzy logic and predictive programming of the heating element would provide the most precise control. Flow meters can tell the computer how much fresh water is entering the heat exchanger and send the appropriate signal to the heating element to respond to the known demand even before the sensors can signal a change. The signal to the elements can be different depending upon the number of groups in operation, the flow rate—single shots verses doubles, whether one steam
SLIDE 11
11 valve or two is in operation and how best to respond to hot water drawn from the tap. The computer can even regulate the auto fill duration and rate to coincide with the element’s ability to maintain temperature. With the correct programming, sensors would be in place as safety devices and for error correction, not as regulatory devices.
(b) Brew Pressure control/Pre-infusion (i) The Purpose of Pressure Profiling
Extraction pressures are another area ripe for debate. We pretty much take it for granted that 9 bars of atmospheric pressure is the gold standard for espresso extraction. While relatively easy to ascertain and probably pretty close to the mark, some coffee respond better to pressures approaching 8 bars as opposed to 9, while the realistic limitations of solenoid valve prevent much experimentation for reliable results north of 10 bar. Early piston driven machines made no mention of the pressures they developed, though they were probably pretty close to what machines are set at today. Noted even less often are the changes in pressures that resulted when the spring in the piston moved from full compression to it resting position. While they were designed to always be under tension to reduce the differential—which is why special care must be taken to rebuild them—there was still some change in pressure over the cycle. Widely considered to be one of the best methods for producing espresso (they are in wide use in Naples, still one of the best places in the world to sample espresso), they are seldom analyzed to figure out what they do better than most modern machines. The single most important thing they did was to pre-infuse the coffee before applying pressure to the filter bed. A by-product of the design, the chamber was forced to preload with hot water from the boiler before the spring could be applied. The steam pressure from the boiler forced the water into the chamber and applied about 15 pounds of pressure to the surface
- f the puck. Once fully saturated with this low pressure infusion—as evidenced by a slow drip
- f espresso from the pour spout—the piston was reset to apply full pressure to the coffee. This
was likely not a design feature, but a lucky byproduct of the design, as it could not be fashioned any other way. Most machine designs from the 1940’s through those in current production make some accommodation to this pre-infusion model. Hydraulic machines from the 50’s used this same profile, as the water pressure above the group simply replaced the spring pressure of the earlier designs. The E61 Faema group incorporates a building pressure regulator monitored be a preset spring to mimic the pre- infusion of piston groups, doing so automatically and without requiring any intervention from the operator. Pre-infusion is designed to allow for higher initial temperatures without extracting the
- coffee. Used in conjunction with temperature profiling, pressure profiling can result in a better
- verall extraction—more even in color and body and much more consistent. Because the filter
bed is allowed to gradually accept the water at lower pressures, it more evenly distributes the water through the puck. As the water seeks the path of least resistance—namely dry coffee that hasn’t expanded from exposure to water—pre-infusion helps insure that extraction pressure is applied evenly throughout the filter bed.
SLIDE 12
12 If the pressures are too high, water is forced too aggressively through the filter bed resulting in a effect called channeling. Some of the espresso begins to extract under pressure while other portions of the filter bed remain dry. Pre-infusion minimizes (or eliminates) channeling, regardless of the quality of or method of tamp. In fact, when Starbucks was looking for a solution to inconsistent extraction between barista—attributed to inconsistent tamp pressure between shifts—engineers set off to Italy to look at alternatives to throwing away thousands of pounds of coffee pre day. At least one manufacturer demonstrated the benefits of pre-infusion by not tamping the coffee at all. Two years later (1995), their existing supplier of espresso equipment filed patent number 5,598,764; a pre-infusion device for saturated group designs. I have never seen one in operation, and given the move to automated machines a few years later, the primary incentive to employ pre-infusion in saturated groups appears to have dried up. The rational given in the patent application, seems compelling. The principle advantages of pre-infusion systems:
- Prevents channeling
- Even saturation
- Expansion of the coffee prior to extraction
- Enhanced extraction when coupled with temperature profiling
(ii) Methods
Several methods are used in modern machines to pre-infuse:
- Air chambers
This is the simplest method of pre-infusion. It provides a very linear pressure curve as water pressurizes the air in a captive chamber in the group. It has proven fairly effective and has the advantage of no moving parts and nothing to service. Simply keep the group clean and there is nothing to maintain.
- Pulsed solenoids
This method is common on some French machines from the early 90’s and is used in many automated machines. While effective when properly setup, the lack of constant, low pressure, could not insure even filter bed saturation as the grind or quantity of coffee varied (singles versus double, for example). Because no other moving parts were added to the system, it was just as reliable as a system without any pre-infusion at all.
- Pump delays
Pump delays are perfect for plumbed in single group machines. Multiple group machines with pump delays are complicated; an early version from San Marco would only allow one group to operate at a time to insure the process was followed. It was very cool to
SLIDE 13
13 watch—as though the machine had a mind of its own—but it made for a slow machine. The Synesso uses a pump delay system of sorts, rather like the early Marzocco GS machines, but the limitations are the same: you can pre-infuse one group at a time or all groups
- simultaneously. This makes for one attentive barista.
The advantage of pump delays is a constant rate of pressure to the grounds with no interruption in flow and virtually no additional parts. If built into the design, no reliability issues would ever come up.
- Spring loaded accumulation chambers
Spring loaded pre-infusion chambers have a slightly different pressure profile, slightly more aggressive than air chambers, they ramp up pressure in response to the resistance of the filter bed and a re less linear, increasing the pressure as the piston extends. A little more predictable perhaps, the addition of the spring, piston and an o-ring results in a little more maintenance, but not much more than changing screens and gaskets.
- Pressure bypass regulators
E61 groups with manual levers provide constant low pressure to the group with much the same effect as a pump delay but with the advantage of operating groups independent of each other. Like air chambers and spring accumulation chambers, E61’s work in conjunction with temperature profiling to allow for hotter water to presoak the grounds before the extraction begins. Because they actually take some of the water away from the process, there is a little more flexibility in the amplitude and duration of the higher temperature profile.
- Accumulation pathways
Cimbali has a slightly different approach that also works quite well. The distance between the brew solenoid and the brew chamber allows for a more gradual increase in pressure and moderates the temperature profile a bit.
(iii) Gicleurs
Gicleurs are often confused with pre-infusion devices because they control the rate of flow though the group. Once the filter bed begins to saturate, however, they do nothing to regulate or moderate the ramp of pressure in the brew chamber. Reducing the size of the jet can solve problem associated with channeling, but can also mask inherent problems with the grind and its affect on flow rates. While they provide for regulated flow in an effort to evenly distribute water over the coffee, they also have another subtle effect on the temperature of the water. This, admittedly, is speculative based upon my observations that groups with jets removed flood the group and result in overheated extractions. I suspect that the gicleur acts in a manner similar to an expansion valve during the initial infusion, reducing the temperature slightly until cooler water from the heat exchanger enters the group. It’s just a theory.
SLIDE 14
14
(c) Moving the heat around and material composition:
Stainless Steel vs. Copper vs. Brass vs. Water
The materials used and the reliance on the ability of different materials to transfer heat energy changes with the design of the machine. We often think of boilers as closed systems with relatively consistent and stable temperatures, but this is not the case. Heat moves through the systems, heating components and stabilizing temperatures using a variety of methods; the conductivity and heat storage capabilities of those materials dictates the design. Water can either move through the system in convection currents where hot water rises to displace cooler water furthest from the heat source, or the conductivity of the metal itself can serve to transfer heat energy. Systems can also rely on a combination of these processes. To get an idea of the effectiveness of different materials to transfer heat energy, note the graph below:
Copper Brass Stainless Steel Water
400 109 17 0.58
Thermal Conductivity of Metals
Expressed as W/(m.K), rate of transfer through each material
The ability of copper to transfer heat energy throughout the system of the machine is virtually unparalleled. Heat is transferred throughout the boiler, not simply by convection of heat energy from the water vapor, but from the conductivity of the metal itself. When working with stainless steel boilers for super automatic machines, conductivity was so poor that even though we expected heat to rise to the upper recesses of the boiler, steam vapor was actually condensing back into water on the interior upper surface of the flange. The steam vapor had to be reheated en route to the valve to produce a dry steam.
SLIDE 15
15 Stainless boilers found in Swiss automatic machines overcome this problem with smaller size and brute force heating elements. Stainless boilers on traditional machines are usually smaller in diameter to minimize the problem of heat transfer, but still use copper for transferring steam to the valves. Even though water is extremely poor in its ability to transfer heat energy in a static column, it has two advantages over metals: It stores significantly larger amounts of heat energy per unit of mass and has the ability to move itself though the system. To take advantage of the thermal mass, the machine has to be designed to utilize this movement of water. Comparative mass values are listed below:
Copper Brass Stainless Steel Water
.39 .377 .456 4.19
Thermal Mass of Materials
(i) System Comparisons 1) Convection based systems
Convection based systems rely upon the placement of the heat exchanger immersed in a jacket of steam or a combination of steam and hot water to conduct heat through a copper
- sleeve. This copper sleeve is what actually heats the water in the heat exchanger. Because the
heat generated by the surface temperatures of the heating element must pass though both water and steam to reach the chamber, the fluctuations in temperature are muted somewhat and produce a more stable environment than that in the boiler itself.
SLIDE 16
16 These systems rely upon the transfer of heat energy to the heat exchanger for temperature control, but also rely upon the transfer of heat energy though the boiler material itself to maintain proper brewing head temperatures. This is why brewing devices for these systems are physically mounted to the boiler itself why direct metal to metal contact. The copper of the boiler is itself a conductor of heat energy to the brewing device. These systems do not, generally rely upon the flow of water to maintain brewing head temperatures. Examples of these systems include: Spaziale, Cimbali, Iberital and Astoria
2) Thermo siphon systems
Thermo siphon based systems rely upon the flow of water as it rises through the heat exchanger to transfer heat energy to the brewing device. The flow rate of the water can be manipulated by restricting or opening the flow through the pipes in the system to customize a particular profile—though few actually do so. The sizing of the pipes determines the dissipation of heat to facilitate flow and plays a factor in the thermal mass of the heat exchanger. The flow rate is unidirectional when at rest, but the flow of water during the brew cycle flows to the head from both the upper and lower tubes, averaging the brew temperature between the two. A cold water injector in the base of the main cavity of the heat exchanger tempers the water feeding the upper pipe. Convection based systems also use cold water injectors, the lengths of which can be varied to customize the duration of the elevated temperature of the machine’s profile. Brewing devices are painstakingly designed to take advantage of flow of water through the system, using thermal mass in a balancing act between thermal stability and recovery from the temperature profile process. Generally, larger heads are more stable while smaller heads are more responsive. It comes down to the kind of profile you want. Machines that use thermo siphon systems include: Faema, Brasilia, Wega, ECM, Astra, Pavoni and the Nuova Simonelli Aurelia.
3) Saturated group systems
Saturated group designs rely upon the convection flow of currents of hot water through a closed system where the brewing device is part of the boiler cavity itself. Because the brewing device is part of the boiler, there is some reliance on conductivity of the metals, but since stainless steel is now common in these systems, the flow of water carries most of the
- burden. The movement of water through the saturated head cannot be regulated and relies
upon the system design to achieve optimal results. Brew water is actually drawn from the base
- f the boiler through a copper tube to balance the temperature.
Some of the newer designs from the United States integrate a small boiler directly into the brewing head design, using PID controllers to individually regulate the water temperature feeding each device. Because the water is heated directly, there is very little reliance on the thermal conductivity of the group material. This system is well suited for bottomless portafliters, where the temperature of the portafilter is less likely to reduce the temperature of the espresso on its way into the cup.
SLIDE 17
17
(d) Tamping
The importance of tamping ground coffee seems to increase the further one gets away from Italy. All of the really cool tampers seem to come from northern Europe and the Pacific Northwest. To be fair, I have a very nice tamper made of stainless steel given to me by Terry Z from Espresso Parts Northwest that I use every day; my criticism of our infatuation with tamping techniques should not imply that they are useless. In fact, I rather like it; even it Italians are puzzled by the obsession.
1) A skeptic’s approach to the value of tamping:
Dosing mechanisms on espresso grinders are inherently inaccurate. The amount of coffee that drops with each pull depends upon the amount of coffee in the chamber as well as the density of the coffee. The more packed the coffee, the greater the density and hence, the greater the weight of the drop. I would routinely tap the side of the doser with my open hand to settle ground coffee to deliver a more consistent drop. Tamping allows the barista to determine, based upon how far the press passes into the portafilter, whether the drop is consistent. The flow rate of water through the coffee is determined by the consistency and even distribution of grind and the density of the filter bed within a known volume. If the grind quality and quantity are consistent, very little variation will be experienced as a result of tamp pressure or technique, especially in pre-infusion based systems. If tamp pressure and technique were fundamentally important to the extraction process, to tamp or not tamp should generate widely varied results; results not experienced in the hands of Luigi Lupi noted above. This position generates the most controversy because many people experience a great deal of variation in the shot quality. The natural tendency is to target tamp pressure or technique as the source of the inconsistency, when grind consistency and volume is a more likely culprit. Machines with large diameter gicleurs that rapidly flood the group can play a factor in channeling and inconsistent extraction. As near as I have been able to determine, there is no glue that bonds the coffee grounds together under pressure, nor is there a locking mechanism formed by the tamping the ground
- coffee. To duplicate the pressure inside the brewing device requires 690 pounds of down
force, a theory we have tested below.
2) Western notions:
It is difficult to look for alternative solutions to extraction variations in a country that names tamping techniques after their inventors. Standards from the SCAA dictate approximately 30 pound of down force to achieve a proper tamp—though in some circles it is
SLIDE 18
18 now 40 pounds. The swift grinder, probably the most accurate dosing device ever fitting to a grinder only tamps to only 5 pounds, though does so throughout the grinding process. This supports the notion that distribution, consistency and quantity are more important than pressure. Tampers come in a variety of shapes, sizes and styles to accommodate virtually any
- philosophy. With the considerable amount of time and effort invested in the process of
tamping the coffee, it is unlikely that any competing, more simplistic solutions will be adopted any time soon. As an experiment into the impact of varying tamp pressures on the extraction process, we conducted the following, relatively un-scientific test. This is not high technology equipment, but simply the machine we have setup in our warehouse using a simple pre- infusion based E61 brewing head and a vibration pump. We devised this simple method to measure the tamp pressure into a ridgeless 14 gram double basket Using the base of a standard 57mm tamp, we applied varying degrees of pressure to the puck and photographed the results. We did not employ any complicated distribution techniques; we simply dropped the portafilter on the counter to settle grounds and removed the excess coffee from the basket to ensure the same volume of coffee for each shot. Essentially, this is how we make coffee here everyday. We made shots of espresso with no tamp whatsoever—simply settling the grounds, a standard 30 pound press, a 150 pound tamp and just to say we did, maxed out the scale at 300 pounds of tamping force. I have never endorsed timing shots as an indication of quality, but the photos reveal an interesting story:
SLIDE 19
19
These shots were pulled with no tamp These shots were pulled with 30 pound Tamp
The color and texture of the crema are virtually identical in these two series, reflecting no discernable difference in extraction as a result of the tamping.
SLIDE 20
20
These shots were pulled with a 150 pound Tamp These shots were pulled with a 300 pound Tamp
For those that simply have to ask, the shot times were virtually identical, though my ability to snap the shot and get the lighting right varied more than the espresso. If there was a difference resulting from tamp pressure, all other factors being equal, something should have happened. And there was no cheating. We didn’t work at this hard enough to fudge the data.
SLIDE 21
21
Section 1.04 Conclusion:
We are naturally curious; its one of the primary forces that drives our industry. It is what inspires us to improve our product, innovate and find resourceful solutions to perplexing
- problems. We should also have a little respect for the talented engineers who have spent
decades refining the art of espresso. The quest should continue, as most of the roasters in the US don’t regularly have lunch with the designers in Italy. But perhaps we should unlearn a few things to get a fresh perspective on how espresso machines work; it is often easier to teach someone to steam milk who has never done it before. The best perspective I have encountered is from Greg Scace, who said that the best way to approach our understanding of espresso machines, is to find a machine that produces a consistent cup of flavorful espresso, start with a clean slate and find out why. Because we don’t have an institutional history of developing and producing espresso machines, we need to start from scratch and perhaps a little humility. What follows is a quotation that seem particularly appropriate, probably the only true thing he ever said: “(t)here are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns -- the ones we don't know we don't know."iv We could probably pay a little more attention to the care and feeding of our grinders, though.
SLIDE 22
22
(i) End Notes
iWBC Procedure for the Measurement of Brewing Water Temperature in Espresso
Coffee Machines
Gregory Scace, Barry Jarrett, Bill Crossland, John Sanders 6.2 Brew Temperature of a Brew Cycle: Specification: The brew temperature shall describe the thermal conditions of water immediately upstream of the simulated coffee cake using two terms, the average brew temperature observed during the brewing cycle, and the one-shot stability. Brew Temperature shall be expressed in degrees plus or minus the stability (for example – 201.5 ± 0.8). In the case of manual data collection, the average brew temperature shall be the temperature observed most often during a specific simulated brew cycle, ignoring temperature observations during the first three seconds of the cycle. Ignoring results during the first three seconds negates the effect of thermometer lag on the result. The one-shot stability shall be one half of the difference between the highest and lowest observed temperatures over the brewing period, negating temperature readings in the first three seconds. For automatic data collection, the average brew temperature may alternatively be the average of all temperature readings during the brew cycle except for those occurring in the first three seconds. The one-shot stability may alternatively be two times the standard deviation of the temperature observations, ignoring observations occurring in the first three seconds.
ii Courtesy of Sean Lennon and Home-Barista.com iii Used with permission, Ken Fox, downloaded from alt.coffee
iv December 01, 2003