Chaff in Coffee: Causes and Solutions

Q: I notice that I am getting a lot of chaff in the coffee when I drop it into the cooling bin when I finish a roast?

A: The simple answer to this problem is Airflow, Airflow, Airflow.  The complication, however, is determining just where the airflow problem is originating: technique, coffee, weather, drum, blower, internal ducting, or external ducting. 

Technique

In roasters that use a single blower for both roasting and cooling- failure to move the damper from the cooling to the roasting positions can cause a reduction in airflow through the drum and may cause chaff to remain in the drum.  Although

Coffee

Let’s start with the coffee.  Changes in coffee, or increases in the use of certain types of coffees can significantly increase the amount of chaff.  Brazils, Sumatras and Ehtiopian Harrars come quickly to mind.  And while the amount of chaff thrown off by any one coffee should not be an issue in and of itself- increases in the use of these softer bean coffees necessitate an increase in the frequency of cleaning and general maintenance of the roaster.  It is pretty simple really: more chaff, more cleaning.

Furthermore, these coffees are the very ones where you will first notice this problem.  They are not unlike the canaries in the coal mine.  Because they produce more chaff you are more likely to begin to notice airflow problems with these particular coffees.  If you see this problem with these coffees and no others, you still have an airflow problem, and you will eventually see it with other coffees if you do not determine the cause and take the appropriate actions.

Weather

Weather is a problem most experienced by the least experienced roasters, or those that have recently moved a roaster or changed an existing duct.  Thermal inversions, abnormally strong or gusty winds or winds from unusual directions can all cause transient airflow problems with roasters.  The best way to determine if you have an issue with weather is by keeping a roasting log and documenting your weather so that you detect any problem patterns.  If you are experiencing transient weather issues there are several things you may be able to do to lessen, or even eliminate, the problem depending on the particulars of that weather, your exhaust set-up and even the surrounding buildings.  And of course, if you determine that your problem was caused by highly unusual weather, you can always not roast when that type of weather reoccurs. Typical permanent fixes include: extending the height if your exhaust, eliminating horizontal exhaust terminations and changing the termination head of your exhaust.

Drum

Over-packing the drum can cause airflow restrictions that will not allow the chaff to be efficiently pulled off the coffee.  Once again, you will see this most often with the types of coffees mentioned above.  Additionally, darker roasts, because of the increase in bean size can cause an over-packing situation, even if the roaster operates fine with that load size at lighter roasts. However, if this problem is appearing after roasting fine for some time, then over-packing is not the cause, although reducing load size can affect a short term solution.

Blower

The heart of drum and air roasters is the blower.  It is the blower that provides circulation that allows for convective action (the dominate form of heat transfer in both air and drum roasters), and that removes smoke and chaff from the roasted coffee.  For drum roasters insufficient cleaning of the roaster blower is the single greatest cause of chaff remaining in coffee.  This is true because drum roaster blowers move dirty air.  This air contains coffee oils, chaff in varying degrees and green coffee dust.  If allowed to accumulate on the blades of the blower these “dirty” elements will begin to reduce the amount of air that the blower can pull through the roaster and eventually can be reduced to a point where there is no longer enough airflow to pull the chaff out of the drum ( this is also the greatest cause of smoke in the roastery).  How often you clean the blower is mostly dependent upon 5 things:

Type of blower (impeller or squirrel cage)

Exhaust run (how much resistance your blower has to overcome to push out the exhaust)

Type of coffees roasted (see above)

Darkness of roast (the darker the roasts the more blower cleaning that is necessary)

Frequency of roasting (the more often you roast, the more often you need to clean)

Internal Ducting

Internal ducting is generally accepted to be the ducting between the drum and the blower, and may, or may not include the chaff collector, if the chaff collector is internal.  External chaff collectors are considered to be part of the exhaust, except where the blower is mounted atop an external chaff collector, many table top roasters are configured in this manner.

It is important that roasters understand how the air moves within their respective roasting system in order to know how, and how often to inspect/clean this ducting.  The frequency of cleaning internal ducting can be quite long, many roasters do this every year or more, but is very dependent upon the cleanliness of blower.  The dirtier you allow your blower to get the likelier you are to experience a partial or complete clogging of your internal duct.  Complete cleaning of this ducting can be a time consuming and back breaking task.  If you are unsure of how this duct runs, the usual frequency and how to clean it- go back to your manual or ask your manufacturer.

External Ducting

Improper ducting materials or poorly laid out external ducting runs are the second greatest cause of poor airflow through restriction; ducting problems are in fact the greatest cause when roasters experience this phenomenon upon initial installation.  Ducts that have too many angles, long horizontal runs, horizontal terminations, restrictive caps, or are of the wrong size can all cause chaff to remain on the coffee by negatively affecting airflow through the drum. These types of ducting restrictions increase back pressure upon the blower, decreasing the amount of air that can be pulled through the drum.  They can often be noted upon start-up of a new, recently moved, or reducted machine.  

However, these types of restrictions are not always noted at start-up, but they will increase the likelihood of trouble in the future, as well as increase the amount of maintenance that must be performed, especially cleaning of the roaster blower.  Roaster ducts should be inspected, if not cleaned, at the same frequency as the main blower.  When cleaned they should be free of oils, chaff and dust. Mushroom caps, or any cap with screen should be avoided at all costs.  China caps should be modified to allow the air to not be pushed back down toward the open duct.  If at all possible all vertical ducting terminations should use a no-loss or low loss stack head.  (see Ask the Expert Roast Oct/Nov 09).  Horizontal terminations should be avoided, if at all possible.

Most recurring problems with roasters come down to airflow issues.  And chaff on coffee is no different.  Perhaps the most important thing to remember about all airflow issues is that a problem in one area can cause connected problems in another, for this reason it is important to try and understand the flow of air; the flow  through your roaster and through your stack.  If not you may find yourself fixing a problem, only to see it reoccur very quickly.

Bean Probes- Types and Placement

Q:

I am intending to add a bean probe to my roaster and I am confused by the different probes, and where to place the probe to most accurately read bean temperature: can you help?

A:

The widespread use of temperature probes in coffee roasters over the last decade or so, even by “old school” or “artisan” roasters has helped make our industry more professional and safer.  Additionally, it has helped spark the movement toward the profiling of coffee by more accurately, and more quickly, measuring changes along the roast time and temperature curve.  In coffee roasting we are generally discussing 3 types of probes: J type thermocouples, K type thermocouples, and RTDs.  Let’s start with the thermocouples.

Thermocouples (Type Js and Ks)

All thermocouples measure temperature in the same way: by utilizing the known temperature gradient of dissimilar metals.  These dissimilar metal wires are housed within the outer sheath of the probe, most often made of stainless steel.  The heating and cooling of these metals create EMF (electromotive force) that is read as voltage this is then translated into a temperature and read by a repeater, controller or computer.  Both J and K thermocouples work in this manner. The practical difference between a J and a K is the range and therefore the accuracy of each type.  A J type thermocouple can measure up to 1380 F, while a K can measure up to 2300 F.  This means that Js are more temperature sensitive than Ks.  Because coffee seldom sees temperatures much above 470F the more discreet readings of the J are a better fit for use as a bean probe.

So, why use type K thermocouples at all.  Type K thermocouples have their use in coffee roasting operations for the reading of temperatures in afterburners.  The higher temperatures required when operating afterburners are at the very limit of what a J can read and continued operation at this level will eventually cause a J probe failure, making Ks much more practical for this operation.  Also, the more discreet readings required in the coffee roasting process are not needed in the operation of pollution control devices.

RTDs (Resistive Temperature Device)

Resistive temperature devices (RTDs) measure temperature differently from thermocouples: RTDs work by measuring temperature induced electrical resistance across the elements.  Once again the elements are housed in a stainless sheath similar to the thermocouples. 

Type J versus RTD

While either a J type thermocouple or an RTD are well suited for use as a bean probe there are definite and distinct differences between the two.  Below is a quick reference list of differences:

RTD

• Able to accurately and easily calibrate
• Easier to bend without damaging
• More accurate
• More stable throughout the profile

J Type Thermocouples

• 2-3 times cheaper than RTDs
• More durable
• Responds faster

The increased accuracy of the RTD comes from the linear nature of the temperature vs. resistance plot, as well as their better stability.  Thermocouples on the other hand are decidedly non-linear leading to more drift from profile to profile, this temperature drift decreases stability.  Generally speaking, most roasters using either a basic set point controller (or using a PID controller in this manner), or a repeater are fine with a J type thermocouple.  While many, if not most profile control systems (and better profiling data-loggers) are using RTDs as these systems are more able to capitalize on the increases in accuracy and stability, thereby justifying the increased cost.

At the end of the day, the decision ultimately lies with the owner/operator, as both of these probe types will work well- so long as whatever is being used to read the temperature can accept the signal from the chosen probe type. And, they are placed so that they can measure the surface temperature of the coffee.

Placement for accuracy

At the risk of sounding too obvious, if you want a probe to read bean temperature, then you need to place it where it is in contact with the beans.  In fact, the probe needs to be immersed in the coffee.  For front opening drum roasters this can be a pretty simple part of the process (see drawings).  For certain types of drum roasters, those where the coffee does not ride up against the faceplate; side openers, bottom openers, etc… then some modification maybe needed in order to accurately measure bean temperature, this modification is often accomplished with a funnel shaped catch that allows a mass of beans to surround the probe.   Fluid bed roasters present a special challenge to the measurement of bean temperature as well.  The larger volume of air, and the movement of the coffee within the chamber, while consistent, will read higher temps than drum roasters; making it difficult for fluid bed and drum roaster operators to exchange accurate roasting information.

For front opening drum roasters the probe should be placed somewhere in the lower quadrant of the uptake side of the drum.  For clockwise rotating drums this would be the lower left, for counterclockwise rotating drums the lower right.  If you need to roast smaller batches then the lower you should place the probe in the relevant quadrant.  The most important thing to understand about positioning a probe for accuracy is immersion, and immersion depth.

Immersion Depth

When attempting to place a probe for reading bean temperature it is important to understand where on the probe (or more accurately how much of the probe) the temperature measurement is actually occurring.  Temperature measuring occurs at the end of the probe back and how much of the end of the probe is determined by the diameter of the probe. You want the probe to be immersed in the coffee to a length of 10 times the diameter of the probe.  So, if you have a probe that is ¼” in diameter it should be surrounded by coffee for the last 2 ½” from the tip back.  Likewise, a 1/8” diameter probe needs a 1 ¼” immersion depth.  These depths can be accomplished bending the probe if there is not enough bean mass depth horizontally (see drawing).

Mounting the probe

The mistake most often made when mounting a probe through the faceplate is not using the correct mounting hardware, or not using any hardware at all.  In picture 3 there are 2 Type J thermocouples both with the correct mounting hardware.  This is a threaded compression fitting, one end is threaded into the faceplate and the nut, with the compression barrel is screwed into the fitting.  The barrel tightens against the probe, holding the probe in position.  If a compression fitting is not used, then the force of the coffee pushing against the probe can spin the probe out of position, possibly damaging the probe.  Roasters often will use tape or some type of adhesive to hold a probe in place, this is not recommended as it will loosen and can damage or even destroy the probe.

When positioning the probe, before bending, and or tightening it is important to ensure that you are not contacting either the faceplate, the drum wall (this will throw off your measurements) or the fins and supports for the drum wall (this can destroy the probe).

Bending the probe

It is often the case, as in picture 2 that a bean probe will need to be bent in order to get an accurate and consistent bean temperature reading.  The probe in picture 2 is an RTD, RTDs are easier to bend and can be bent using the thumb as a rest, and bending over the pad of the thumb.  It is more difficult however to bend a Type J thermocouple and care must be taken not to crimp or damage the interior of the probe.  It is best when bending a J thermocouple to use a mandrel (a pulley or rod will suffice) that is at least 21/2 times the diameter of the probe.  Do not attempt to bend a thermocouple without using something to bend around.  With both probes you need to take care not to crimp the probe when bending for a 90 degree angle.

More Accurate Tools, Safer Operation and Better Coffee

Regardless of which type probe you chose, or your reason for choosing it, using a bean probe to help you monitor your coffee during roasting will be one of the best business decisions you will ever make.