Want to build an automated HLT: advise?

I have started looking into building a automated hot liquor tank for heating mash water unattended or remotely. I have a few ideas and would welcome advise from the community.

Here is a verbal sketch of my prospective build:

I would like to fit a scrap 15.5 gallon keg with 2 ports for water in and out. One to fill and another to drain. These would be controlled by solenoid valves.

There would be an electric hot water heater element installed in the vessel to heat the water.

I would have something like these 2 sensors:
Precision Epoxy Thermistor
https://www.adafruit.com/products/381

eTape Liquid Level Sensor
https://www.adafruit.com/products/464

to measure the temperature and water level.

I was thinking that the whole thing could be controlled by an Aurdrino micro-controller which in turn could interface with some custom python control programs that I plan to write.

An imagined use case would look something like this:

Saturday evening I set the system to heat 4.4 (or whatever) gallons of water to 165 degrees F at 7am the next morning. At 7am the computer wakes up and tells the controller to open the fill valve. When the water level sensor reaches the designated 4.4 gallon mark, the valve is closed and the heating element is turned on. When the water is at 165 according to the temp probe the heating element turns off. The second solenoid valve opens and gravity drains the hot water into my cooler mash tun.

In the mean time I get up and pour a cup of coffee and walk out the the garage and dough in.

My main reason for doing this is to improve my programming skills and begin learning about micro-controllers. I am aware that this may be an ambitious project for a beginner but I find that learning by doing is the best way for me. I have about 2 1/2 months to work on this relatively uninterrupted and don't expect to have it finished by then. I am also tight on cash and want to do as much scavenge and scrap as I can to get to an end project.

Any advise from seasoned builders or opinionated loudmouths would be great. I am in the early planning stages.

Thanks a ton(tun?).

Well, I may have put this in the wrong section. Should it go in Automation?

Anyway, looks like the eTape sensor is only good up to 150deg F so I am looking into other options. I like the look of some ultrasonic sensors.

I also found some cheap Ardruino clones at this site: http://www.dealextreme.com/ They are well reviewed and cost a few buck less than the ones on the mainstream sites.

Thanks again to anyone with ideas of advise who wants to share.

If I were you I'd use the 'clunky' float switches. Two of them: 1) Max fill level so you don't over float and 2) Min liquid level for the heating coil to be submerged but still running (so you don't burn it out by accident with draining).

Get some with 1/2" MPT threads and they will screw right into your 1/2" NPT ports (or weldless fittings) and give you access to their wires outside your HLT.

edit = MLT <> HLT

I think I saw someone here using a tube (like sight tube) and proximity sensor of some sort. I got to start bookmarking all these methods...

gx1400, thanks, that is a good safety feature. Don't want to burn down my garage. Wife would be pissed.

RighBenn, I think that you are talking about 'bubbler' pressure sensors. After doing some more reading, this seems to be a common solution.

Right now I am tending toward using an ultrasonic sensor to gauge the liquid level. They are cheap and some folks give good reports on them. For example: https://www.homebrewtalk.com/f11/ultrasonic-liquid-level-scored-off-ebay-226760/

the biggest issue you will have to overcome will be thermal carryover. if you turn the heating element off at 167 it's basically got "momentum" and will keep heating to a certain extent. this is why PIDs are so awesome. they are always monitoring how fast the temp rises for a given amount of power to the heat source and ramps down as it approaches the temp to not overshoot.

what's the difference in programming the rig to fill itself with water at 7am instead of filling it the night before? this would eliminate a need to constantly check volume. i definitely agree you want a float switch so the element can never be on if the water level is below the heat element.

slakwhere said:

the biggest issue you will have to overcome will be thermal carryover. if you turn the heating element off at 167 it's basically got "momentum" and will keep heating to a certain extent.

Click to expand...

the 'thermal carryover' is not an issue for a standard HLT. it only comes into play when using a HLT as a HERMS heater. once you cut power to the heating element, there are only a few dozen residual BTUs (if that) in the metal of the element.



to simplify it, i would install maybe 4 or 5 small liquid sensors up the side of the tank. have the bottom one positioned where the minimum fill for the elements to be covered is. have one at the top for a MAX (or better yet, have a float valve at the top so that even if your electric valve or arduino fails, it still cant overflow).

then install maybe 2-3 other liquid level switches at specific volumes in between; maybe at the 3, 5 and 7 gallon marks. you wont be able to specify "4.2 gallons", but you could just tell the arduino to stop filling when the 5 gallon sensor was tripped. i would use small screw in sensors like
http://www.fluidswitch.com/pages/OpticalLiquidLevelSensorOS-900.htm

slakwhere, the auto-fill feature is not strictly necessarily, you are correct. I want in 1) because it is cool, and 2) so that I can initiate the process remotely, via email, SMS, or twitter.

It may make sense to relegate that feature to a phase two list, it is true.


audger, what would be the advantage of those over a single sensor setup? Also, are you familiar with those type of sensors? They are a bit over my budget. Do you think a cheaper version would work?

I made an order of an Arduino Uno, a cheap temp sensor and a ultrasonic distance sensor. The distance sensor only cost about 5 bucks, so I am going to see if I can get anything meaningful out of if if I point it at slightly turbulent water.

Can anyone tell me about solenoid valves? Are they a good choice for this application? Will the closed by default valves reliably fail closed?

You guys are great, BTW.

As a former automation engineer, I highly recommend some redundancies, failsafes. I can see a solenoid or sensor failing an your garage flooding.

I would include a maximum fill time. I would also use dual solenoids and perhaps a flood sensor.

audger, what would be the advantage of those over a single sensor setup?

Click to expand...

unless you can find a single sensor that works with 212+ degree liquid that spans the entire 1-2 foot range of liquid height, i dont know if you have many other options. the code would certainly be shorter. ultrasonics will drift as the air temperature and humidity increases, and (especially cheap ones) arent the best at accurately measuring liquid surfaces, especially with bubbles. this is in general, i dont know what specific ones you are looking at. at the very least you will want a mechanical float valve at the max height mark to positively stop water from flowing.

you could also probably get an accurate water pressure sensor and install it at the very bottom of the kettle. the pressure would be fairly linear, and its simple math to do the conversion to gallons.

yes, normally-closed solenoids usually fail closed, unless something physically gets stuck inside them. solenoids in general are 'ok', they give a large water hammer shock when snapping closed that you need to be mindful of (particularly when using plastic plumbing). motorized ball valves are easier on the plumbing but are a few extra bucks.