Friday, 13 April 2012

Heat Exchanger Temperature Surfing Research

This is a follow up to my earlier post "My UKBC Prep And Practice Area" and the brief discussion in the comments area below it. It's about traditional barista skills rather than about the UKBC, and it's a little long and probably quite boring for many readers. You have been warned :)

Times have changed
Espresso technology has moved on. World Barista Championships machines (and individual national/regional events) are run on machines with very stable temperatures as a result of using multiple boilers combined with PID digital temperature control on each brew head. In this working environment baristas tend not to worry about achieving a particular temperature profile. (Perhaps this is why pressure profiling has become more of a consideration.)

But I own a Heat Exchanger machine (a commercial Faema Enova), which is now considered old school amongst many well established baristas. The upshot is that I DO need to worry about temperature profiles, and need to be able to control my brew temperature using manual techniques. I am the PID!

I could of course just sell my HX machine and upgrade to more modern technology, but a big part of me thinks it would be a cop out.  I see learning how to get the best out of a HX machine as part of my apprenticeship as a barista, so that I have a stronger foundation and understanding if/when I do eventually upgrade. Some would say I'm foolish for taking this approach... and I might even agree. Maybe it's a trait I've inherited from my father, but I've never been one to take the line of least resistance. Good things shouldn't come easily.  Also, I feel there's something in my approach that is respectful to those who have gone before me over the years, who came up with theories, developed ideas into practical processes and solutions, and ultimately paved the way for today's technology... for example:

and although his was a slightly different field, this gentleman, who sadly passed away recently at the grand age of 90 --->

Getting back to Temperature Surfing
Anyway, I've talked about temperature surfing for a long time on, and performed it to an adequate degree on my single boiler (non-HX) Gaggia Classic, but as I researched more recently, and due to helpful pointers from the comments in the aforementioned blogpost, I came to realise that I didn't actually know very much about how HX machines work, how their design influences brew temperature, and how to control the temperature to produce, on demand, different temperature profiles throughout the shot.

I've now read more, and walked the google trail, which led to hacking my way with a machete through the forums jungle from thread to thread spanning several years of posts1 .  I now feel a little better equiped to use a HX machine in a controlled and accurate manner.  Here are some of my findings. As always there remains a lot more to learn, and once again I'm not saying anything new. I'm just documenting my learning path.

So... HX machines have a temperature hump. When not in use the water in the exchanger heats up and circulates around the brew head, resulting in brew water temperature becoming too high. This is visibly evidenced by steam and spitting when the brew button is pressed.

My machine is built to be powered by a three phase electricity supply (415V), but has been rewired to operate on a single phase supply (240V), which is the standard voltage in the UK. This began to worry me recently as I wondered whether it might affect temperature stability, and might make the hump more pronounced. By this I meant that at the start of the shot the brew water temperatre would be low, rising to a peak in the centre of the shot time, and tailing off to a low temperature at the end of the shot.  After exploring the design of the HX system and the E61 group head I'm now no longer worried. I'll explain why, but first let me show something that contributed to my concerns.

The "HX Love" article on Home Barista (see notes below) discusses the humped profile and profides the following diagram.
Figure 1. Temperature Profile of a commercial HX machine

The author, Dan Kehn, also posted the following video.

So I proceeded to track my own machine's hump. This is with an empty basket and a single spout. (My video is obviously in degrees C rather than F)

Here are the temperature profile results of several logged flushes.
Figure 2. Temperature Profile of my commercial HX machine.
First off, there are a few mitigating factors to consider here.
1. My temperature probe is inexpensive, and therefore not fast acting. I estimate a 2 second delay between probe reading and display.
2. There is also approx a 1 second delay between my pressing start on the timer and hitting the brew button.
3. Whilst the probe is rising in temperarure (playing catch up), the water temperature is actually coming down... even though the chart shows it rising.
(The important point is that the chart shows the point at which the water temperature peaks.)

The point however is that I could not understand why my temperature tailed off to around 75C in 30 seconds when Dan Kehn's seemed to stay fairly high. This baffled me. Actually I do still wonder about it, but not much because I now realise I was looking at this all wrong.

The profile shown in Figure 1 is not comparable/compatible with the two videos beneath it and with my own temperature profile chart. It must relate to the temperature DURING BREWING... i.e. with a full portafilter basket! No wonder I could not replicate it. (I feel stupid for my mistake, but can also see how easy it is for someone to make that mistake.) The throughput of water from an empty basket is perhaps six times greater/faster than when the basket is full. That is, a full basket may take 30 seconds to fill a 2oz glass, whereas an empty one will only take perhaps 5 seconds.

Not having taken the time to properly understand this scenario, I had thought my machine's temperature stability was extremely poor, and begun to wonder if the single phase conversion had caused this.

I have now learned more about the HX system. The brew water comes from the circulating HX system. Operating single phase will not affect temperature stability because the temperature is regulated by the temperature of the water in the boiler, which remains relatively constant as no water goes into or comes out of it when brewing espresso... only when running water through the hot water dispenser, or venting steam from the wand.  Inbound water directly enters the circulating HX system and mixes with the water already in there, replacing outbound brew water.

So what?
Well, first off I've used the chart in Figure 2 to provide an indication of how long my flush needs to be in order to achieve a fairly flat brew temperature profile.  At around 9 seconds the hump has reached its peak. The temperature has maxed out and the brew water is no longer steam, and is no longer spitting. Adjusting for the 2-3 second delay mentioned above, I now know that a 6 second flush will get my machine to 92C.

Secondly the brew temperature will stay at approx 92C for 7 seconds when flushing with an empty basket. The water throughput during this time is approximately 3 oz, which is of course sufficient to produce a double shot of espresso (taking into account water retained by the puck during expansion of the grinds).

Thirdly, after this point the brew water temperature would slowly tail off, but that is irrelevant as brewing would have ended by then and the HX system continued reheating the circulating water.

Fouthly, if I want a hotter or cooler brew temperature (within a degree or two) I have options. Reducing or increasing the flush time is one way to do it on the fly, and I can use Figure 2 to indicate how many seconds to reduce/add to the flush.
If I want the machine to run at a higher or lower temperature more permanently (to suit a particular batch of beans, for example), I can adjust the pressurestat. This will increase/decrease the boiler water temperature, which will in turn increase the HX brew water temperature. (If I do this then I presume I will need to rechart the temperature profile of the flush.)

I am still unsure as to how long I should wait between the flush and inserting the portafilter/pressing the brew button (i.e. the rebound time), and plan to continue researching and experimenting with my machine to see what effect different rebound times have.

Some conclusions
Oh, I dunno! I learned some stuff. I have more to learn.
I guess one conclusion is that is not just for home baristas.  Most of the things I've learned in this session have come from that website/forum and are completely relevant to burgeoning profession baristas such as myself. The folks who posted such amazingly informative articles on there are certainly not amateurs. We find learning wherever we can get it, and hopefully in writing this blog I will help someone else to gain access to learning resources more easily. Also, if anyone reading this finds that I've misunderstood or missed something then please feel free to post a comment prodding me in the right direction. Self-education is prone to error!!

1. Here are some pieces of the trail: