HERMS (Heat Exchange Recirculating Mash) Automated Brewing

Primary tabs

Project Update - August 6, 2013 - Update 1: Filtering

One of the finished brews done prior to chlorine filter installation was ready to keg, but the off flavors were strong enough to discourage it.  The jury is still out as to the cause of the off flavors, but in an attempt to save the beer, and to play around, we acquired a 0.5-7 micron beer filter.


The filter sits between two kegs, one with beer, the other to collect the beer after the filter.  The filter has takes disposable inserts in the shape of discs, and comes in several sizes.  I tried each size on the beer with off flavors, which was also very cloudy.  This beer didn't use any fining agents, so it was a great test for the filter.  I tried the 5 and 2 micron filters separately (rough and polish), but the cloudness remained.  After trying the 0.5 micron "steril" filter though, the results were impressive. 

It did seem to help some with the off flavor, but not enough to salvage the beer.  If anyone intends to use this filter, I recommend running at least a gallon of water through it before use.  The filter imparts a kind of gram cracker taste for the first gallon or so, and not in a good way.

After opening the filter up, you can see a lot of still suspended yeast.


All in all, a good tool to have in the toolbox, but I will probably use it sparingly, as the filters add about 4$ a gallon and an extra step.

Project Update - July 26, 2013 - Update 3: Yeast

We're talking about yeast.

Yeast is one of the most important components used to make a good beer.  Each strain of yeast has it's own charactaristics, some produce fruity esters, and prefer higher fermentation temperatures.  Some like it cold, and produce crisp clear beers.  Some ferment more of the complix sugars producing dry beer, and some leave it sweet.  There is clearly a lot to learn here, so we are ramping up to try to understand yeast, and to experiment.

Fortunately, Andrew just happens to be a bio-chemist, and is helping us gather all the tools we need to study yeast in relation to beer, and to experiment with it.

So far, we are preparing to:

  1. Do yeast cell counts
  2. Wash and maintain a store of yeast
  3. Freeze and store yeast long term
  4. Harvest and select wild yeast, to discover our own strain
  5. Detect bacterial infection of the yeast and beer

For the yeast cell counts, we have procured a microscope (Thanks J.R.), a Hemocytometer, and we are working to get some trypan blue.  The hemocytometer is just a slide for the scope with a chamber with a grid.  With a microscope, you can count how many yeast cells are in each cell of the grid.  That gives you a total count of yeast per volume.  The trypan blue is a dye that will only color dead cells.  This lets us easily ignore any dead yest from our count.  We typically want about 200 billion cells for every 5 gallons of beer.

We also want to be able to grow yeast strains from samples taken from various places.  You can re-use yeast from previous batches, grow yeast found in other beer, or even find a new wild yeast and just see what happens.  It is also very important for yeast to rehydrate and let it multiply in a starter before pitching it.  To do all of this, we're working on building a temperature controller for the fridge in the basement, and a stir plate.

We have built a stir plate from one of the many online DIY stirplate guides.  The case was made in sketchup and 3D printed, and the components / bread board fit right in.  A stir plate is used to keep the starter moving, preventing the yeast from settling and going dormant.  It's a great way to get a higher cell count for pitching, and another tool to ensure a great start to your beer. 

The temperature controller components have all arrived.  We are looking to try to control both the freezer and refrigerator at separate temperatures using an MSP430 microcontroller.  This will allows us to brew beer in the larger fridge space, while storing our yeast at safe temperatures.  There is a great article on homebrewtalk.com describing the process.

Finally, if you have ever ruined a beer through bacterial infection, wouldn't it be great to know your home grown yeast was infected in advance of pitching?


Project Update - July 26, 2013 - Update 2: Mash

The refractometer issue has opened up some discussion about verifying and studying our mash schedule.  We are doing a stepped mash, and the temperature change at each step has an effect on the body of our final beer.  It would be really great to verify that the steps we schedule are doing what we expect.  One can taste the difference in a full or light bodied beer, so we are brewing beers from identical ingredient, while making changes to the mash schedule.  At various points of the mash, we are gong to do an iodine test.  This will help us estimate how much remaining starch needs to be converted to sugar.  As the mash progresses, we can record when the starch test begins to show similar results to the previous test, indicating starch has been fully converted.

I was able to order some Iodine 2% from amazon for about 10$.

If it turns out that starch is fully converted well before the last step in the mash is complete, or another variant, we could better time our mash steps.

Project Update - July 26, 2013 - Update 1: Refractometer

Our current brews are almost done fermenting.  We're just slowly bringing the temperature down each day by a few degrees to bring the yeast to rest, and to allow the beer to clarify some.  Andrew's brew tastes delicous, the stout is also promising, but the two ale are a little off.  There lots of reasons for a beer to taste a little funny, but in this particular case, it seems the chlorine content of the water didn't burn off in the pre-boil as hoped.  If we consult Palmer, he explains it pretty well.  The beer may also come back if we give it more time in the secondary.  Meanwhile, we're addressing the chlorine problem with this:


This is a high flow, whole house carbon filter.  It does up to 30,000 gallons before it needs to be changed, and filters down to 25 micron.    This is enough to remove the chlorine from the water, but leave much of the rest of the water profile in the default range for our location.  Andrew has found that New Haven water is also typically low on calcium.  Adding gypsum should solve this, and should be available from Luck and Levity around the corner. 

We started using a new device for measuring the gravity of beer, a refractometer, and have been taking readings of our brews every couple days while waiting for fermentation to finish.  Levels of sugar in a beer is referred to as Gravity.  You take a measurement for this before adding yeast to the beer to see how much total sugar is present.  Later, after the yeast has done it's job, you do another measurement.  By subtracting the second measurement from the first, you can determine how much sugar has been converted to alcohol by the yeast.  Typically people use a hydrometer for the purpose, but a refractometer is an alternative, and is really convenient, needing only a few drops for the test. 

We learned the hard way that the refractometer may be reliable for the original measurement (Original Gravity or OG), but after fermentation, an additional calculation is required.  The refractometer was only reading that half our sugars were fermented.  This is referred to as "stuck fermentation", and something must be done to get it going again.  We added more yeast to each fermenter, but after a few days, there was no change.  We began to wonder if our mash schedule/equipment was inaccurate, and we created too many unfermentable sugars.  Finally, after trying the hydrometer, we discovered the refractometer to be inacurate, and that it needs to be adjusted when measuring the final gravity.  Northern brewer has a great calculator for making the adjustment, and it shows that our beers are completely fermented, and ready to clear and keg.

Crisis averted.



Project Update - July 19, 2013

The machine is up and running, and we're making beer.  I thought it would be nice to finally get some pics of the finished product here, of when it's brewing, and some of the challenges left to solve.

First, this is what she looks like.  Each pot is 20.5 gallon capacity, HLT on the top left, MASH in the middle, Boil Kettle on the right.

The electrical panel is based on the open source BrewTroller software, with the OpenTroller DX1 controller.  There also various other components, controls, relays to drive 2-way and 3-way valves, pumps, heating elements, volume sensors, and temperature sensors.

This pic shows many of the valves, electrical, and water-in/out utilities.  The machine will control all fluid movement for the brewing process, including the initial fill, and cleaning of the machine. 

This is the Mash Manifold.  The 3-way valves are the blue valves, each can optionally move fluids in one of two different directions.  The two shown here will move fluids directly back into the mash, or through the coil in the HLT, or over to the Boil Kettle if mashing has completed.


Here are some pics of the machine in action.

This is the HLT after it's been auto-filled to 10 gallons.  You can see it heating up, getting ready to transfer heat to any wort that is pumped through the coil.

This is one of several level sensors, on the other end of this tube is a pressure sensor similar to one you might have in a digital scale in  your bathroom.  The more water in the pot, the higher the pressure sensor reading.  This is how auto fill and transfer of just the right amount of fluids is controlled through the brewing process.


After the HLT is heated and ready to go, some of the fluids are transferred to the MASH.  This is the mash filling up:

After the water is at the right temperature, the grains are added.  The temperature of the water already takes into consideration the temperature of the gains such that when mixed, just the right starting temperature is achieved.

Recirculation usually takes around 90 minutes, depending on the recipe.  You can see the wort getting darker as more sugars are extracted.

After the mash, the wort is slowly fly sparged to the boil kettle.  The machine also supports various other kinds of batch sparging.

As the fluids are boiling, hops, sugars, and spices can optionally be added.  The machine will alert as these steps need to take place according to the recipe.

The counter flow chiller is a stainless steel tube-inside-a-tube design.  Wort flows in the inner tube, water flows in the outer tube.  As the boil comes within 20 minutes of completion, a pump will start moving hot fluds through the counter flow chiller to sterilize it.  As the boil completes, cold water is pushed in the opposite direction through the counter flow chiller which will bring the finished wort in the boil down to a temperature which the yeast will enjoy without killing them.

Another purpose of the recirculation is to get the fluids in the boil kettle moving in a whirlpooling motion.  This causes all the particulate, which is lighter than the wort, to pile up in the middle of the boil kettle.  Drawing the finished wort off the side of the kettle will leave most of this unwated particulate in the kettle and not in the fermenter.


This is the effect of a good whirlpool.


Making a stout for the first time on the machine was really interesting.  You can see the stout in the high temp food grade silicon tube distinctly.

Here's what the stout looks like before boiling, quite dark.


After the boil, I took a little stout, cooled it, and pitched my yeast to get it started as the instructions suggest, in only a few minutes, the dark liquid was a cloudy brown with yeast.  The paper towel was dipped in Star-San, a food safe sanitizer, and is helping me keep the bugs out.

Fermentation is being done in a freezer controlled by a ranco temperature controller.  These brews should be completely finished fermenting in about a week.

Stay tuned for more updates, stop by MakeHaven on a Tuesday if you want to try one of our first beers as we continue to refine the process.



Project Update - July 18, 2013

Wow.  A lot has happened since the proposal late last year.  We're brewing, and it's really fun.  For the build of the machine, we realized virtually every functional goal we set out for. I'll try to squish everything into a project update.

Soon after proposing the project late last year, I was approached by Karen Bliss, Professor of Mathmatics, at a Tuesday open house.  Karen was looking for automated and reliable collection of various data points in the brewing process.  The idea was to apply for a grant from the Reintegrate New Haven project sponsored by the Arts Council of Greater New Haven.  The focus of the grant is to "encourage exciting possibilities of collaborations, ... that fosters relationships and dialog between the scientific and artistic communities in the region" - http://reintegratenewhaven.com.  She was applying for the grant with her colleague Charmaine Banach, Professor of Interactive Digital Design.  We quickly agreed to collaborate, and Karen and Charmaine set out secure the Reintegrate grant.

Out of dozens of appicants, our project was one of the few selected, and our project began in earnest.  We were joined by MakeHaven members Michael Mckien, and Andrew Chastain, Quinnipiac Math students Katelyn Stoll and Caitlyn Hannum, and Design students Nicholas Baldoni, Tim Ostheimer, Samantha Epstein, Nicole Cornetto, Lawrence D’Onofrio, Ricky Lencsak, and Zach Freed.

Countless hours and contribution by all invoved made a very fun, interesting, educational, and rewarding project.

Check out all the ups and downs on the Reintegrate blog below with lots of fun details and updates for the project:


You can see the machine a little more at some of the recent MakeHaven publications:



As far as we are, there is still more to do. 

Sign up if you want to brew, learn about the machine, improve it, or offer support in advice.


September 15, 2012

The Colonizer (Yeast)

It's time to build a fully automated brewing system at MakeHaven.

This is a really fun project, needing a variety of contribution; design, metal working / welding, chemistry, electronics, software development, and cost-saving creative thinking.  There are countless approaches to building automated home brewing solutions.  From 5 gallon, single vessel, dishwasher sized solutions, to 20 gallon, three vessel builds.

We are aiming for the more ambitious of options, because why not.

The most popular and reliable manual build looks a lot like this:



In summary: One vessel is where you add the malt grains to extract sugars, another has ~160 degree water and is used as a heat exchange, and the last is where you boil while adding flavoring.  It has a device to cool the beer after the boil, and a couple motors to move liquids around supporting various processes.

We will build this, but start to finish automation.

Version 1.0 will:

1.  Fill tanks with water to precise level, without intervention, using water level detection with pressure sensors.

2.  Step Mash - Change the temperature while extracting sugars from the malt grains on a predetermined schedule.

3.  Support two batches concurrently (in separate stages), each between 5-20 gallons at ~2.5 hour total run time.

5.  Fully electric, with stainless steel safety float switches to ensure elements never dry fire.

6.  Up to 10 audio/visual alarms for hop additions during boil.

7.  Remote monitoring over web / phone.

8.  Automated cleaning cycle.

And more...  Most of this functionality is already supported on the open source BrewTroller project.  Here is their electrical/plumbing illustration and description.

We have begun to put together a parts / price list.  We would like to reduce costs though salvaging and building in-house wherever practical, while still tracking and considering "Ideal" options.

Here is a preliminary parts list. Among the missing from the list are metal framing for the base stand, and electrical housing for the main electronics.  We're hoping these are feasible to roll our own.

We want to have a functioning brewing system in two months, to be fully prepared for the holiday season, so lets take two weeks to see if any materials can be scavenged up before putting orders out.

The completed build may or may not live at MakeHaven, depending on feasibility, so we can relocate somewhere local as long as the system is accessible by members.

If you want to contribute to the build, have any of the parts, are willing to help purchase parts, can help research better deals for specific components, want to think up crazy cost saving ideas, or just wanna help plug stuff in to make it work, reply to sign up.

Lets get started.