Propane throttling?

[rpierce]

I'm building a custom designed HERMS system from scratch. I am using low pressure (11" W.C.) propane, Honeywell valves with standing pilot lights, 1/2" NPT pipe, and 90-100k BTU jet burners.

I'd like to make the boil more repeatable by using a level sensor in the boil kettle and controlling the rate of boil via computer. Having the computer turn burners on and off via the Honeywell valve works great for holding temperatures, but I need a way for the computer to control the heat output of the burner. So I need some kind of actuated valve rated for that amount of BTUs that can throttle the amount of propane. I figure a ball valve probably isn't a good choice. Maybe I need a needle valve? I'm using an Arduino Mega for the low level control, so I'm flexible; I could use servo pulses, drive a stepper motor, or possibly output an analog voltage signal.

Any thoughts?

 

[MrNatural]

My thoughts are: good luck.

You can't use temp because gentle boiling wort and boiling over wort are the same temp.

How could you use a level sensor and compensate for boil-off? I would think you'd have a bigger mess boilover-wise and programming-wise versus just finding the sweet spot and let it do its thing consistently while keeping an eye on it.

Seems pretty likely you'd just end up pulsing under/over the sweet spot anyway, even if you did go through the efforts of engineering and programming successfully.

Even the electric folks, with simpler heat control, find the sweet spot and let it be.

Anyway, good luck. Interested in seeing what you come up with.

 

[joeycmsu]

I think you could make a stepper motor on a needle valve work. The valve only needs a Cv of around 0.11 for 40SCFH at 11"wc (100kBTU at 2500 BTU/scf for propane).

I don't know how you would use a level sensor, but I think you could make it work on temperature control of the kettle. That is what I have experience with in commercial breweries so I can't see why it wouldn't work in home brew.

If you really wanted to control to heat output, you could get the valve curve for whatever valve you choose and calculate expected gas flow for different valve openings, then interpolate in your program to be able to set a specific BTU output. I would just use a temperature PID controller though.

-Joey

 

[rpierce]

My plan is to measure liquid volume and compare it against a projected ramp of where the volume should be at X minutes into the boil so that I end up with Y gallons of wort. I'm going to use a dip tube with an acquarium pump and a pressure transducer as a level sensor, which I hope will give reasonable volume readings. If I know how fast I'm boiling off relative to how fast I should be boiling off, I can use that as input to a software PID controller that will control the propane to keep the boil on target.

Even if I end up adjusting it by hand, I will still need a throttle valve. I've got access to a machine shop and could probably couple a stepper to a regular needle valve to make it computer actuated. But gas fittings are something I do not know much about and could really use advice here. Given these specs, any suggestion on a needle valve to purchase? I'm guessing a lot of people use low pressure propane and need a way to throttle it....

 

[joeycmsu]

Something like a swagelok or Parker needle valve would work, it doesn't need to be anything fancy since you're on the low pressure side of the LP regulator. Brass, bronze, stainless, or carbon steel would all work as materials of construction.

Here is a typical needle valve, you might find something easier to mount a stepper motor to:
http://www.swagelok.com/search/find...N4&item=42e66d92-14ca-497c-9246-d6922ca2a4c3#

Here's the data sheet:
http://www.swagelok.com/downloads/webcatalogs/en/MS-01-85.pdf

Swagelok stuff is kinda pricey, but you could get something similar from Parker, Watts, Hoke, etc.

 

[rpierce]

Thanks. But that needle valve was 1/4" NPT and the largest they listed for that valve style is 3/8". All the parts in the system are 1/2" NPT which I thought was necessary due to the high flow rate required to get 100k BTU.

 

[MrNatural]

My plan is to measure liquid volume and compare it against a projected ramp of where the volume should be at X minutes into the boil so that I end up with Y gallons of wort.

This is where I get a brain cramp, and hoping a documentarian or two might have some input on the effects of environmental variables on boiloff.

I would think the "projected ramp" is not a constant, but is influenced by ambient temperature, barometric pressure, relative humidity and wort specific gravity.
NASA could probably do the compensation algorithms, but I feel further onset of cramping.

Just to be sure I understand your quest, you want to tailor your boiloff rate to arrive at desired final volume at end of recipe boil time (60/90 min)?

Bunch of needle valves here:
http://bigway.com/needle-valves-c-118_133/

 

[rpierce]

Yes, I start with X gallons wort, I want to end with Y gallons of wort, and do so in Z minutes of boil. So I am guessing that I can change the power output to change the boiloff rate. There's going to be an upper limit where the kettle would boil over, and a lower limit where it stops boiling, but I'm guessing there is a range where I can control BTUs to the kettle to get a variable boiloff rate, and use PID control to steer it towards the target value.

 

[rpierce]

I'm not sure I'm following the math here. I see that 1 cubic foot per hour is 2500 BTU, so I follow that I'd need 40 cubic feet per hour to run the burner at full throttle. That's 2/3 cubic feet per minute, which is almost 6 gallons per minute.

Looking on eBay, I see a Parker 8F-V12LR-SS needle valve that is 1/2" NPT. But I see it is only rated for 1.23 gpm. So am I correct that, wide open, that valve could only handle less than 25k BTU? Or is the valve flow rated for liquid, and more gas can flow through the same valve?

 

[joeycmsu]

1 standard cubic foot of propane has 2500 BTU's heating value. A 'standard' CF is at atmospheric pressure and standard temperature.

The propane in the bottle is stored in liquid form. Propane has a density of 0.1175 lb/cubic foot, so there are 8.5 SCF in 1 lb of Propane, or 170 SCF in a 20lb bottle.

If you run your burner at 100kBTU it will use 40 SCF/hour. That gets back to the 0.11 required Cv, which is around a 1/4" needle valve. You can use a bigger valve, you'll just end up keeping it in the 10-25% open range.

Here is a data sheet for that Parker valve that lists the valve coefficients
http://www.parker.com/literature/In...Catalogs/CAT_4110_NV_NeedleValves_Dec2010.pdf

I did the calculation again with more information (specifically for LP), and it gave a required Cv of 0.34. The parker V6 series valves would give you the correct size valve orifice, or you can use the V8 like you found and it will just be a little too big, which won't be a huge deal.

 

[mash-mongrel]

...............

 

[rpierce]

I'm using a Honeywell VR8200A2132 controller for each burner. This provides an on/off valve for the burner that responds to a 24VAC call for heat, but it also provides gas for a pilot light and has a thermocouple. So if the pilot goes out, it will turn off the gas to the pilot and ignore calls for heat. This should be safe if the pilot light goes out, seeing as this is generally what furnaces use.

What I'm talking about here is a valve that will be installed after the Honeywell controller that can throttle the amount of gas sent to the burner during a call for heat. I figure for my HLT and MT, I don't need one, as I can cycle them on and off to hold temperature. But I'm going to want my boil burner on constantly with a variable amount of power during the boil.

 

[ThreeDogsNE]

Thanks. But that needle valve was 1/4" NPT and the largest they listed for that valve style is 3/8". All the parts in the system are 1/2" NPT which I thought was necessary due to the high flow rate required to get 100k BTU.

How much are you trying to boil? My gas manifold is 1" pipe supplying 4) 12" 3 ring stock pot range burners. I used the valves that Imperial uses for their ranges, which are 1/8" pipe, with one valve for the outer ring and one for the inner two rings. The orifices on these are 46 wire gauge, IIRC. I have 1/2" branches from the manifold, then 1/2" to 1/8" reducers just below the valves. I have Megapot type 20" diameter kettles. I get a strong rolling boil on ~14 gallons in them, and end up throtting back the valves, and still have at least 1.5 gallon/hour boiloff. If you are using 1 valve, 1/4" input would still give you twice the gas capacity I have.

 

[crane]

Yes, I start with X gallons wort, I want to end with Y gallons of wort, and do so in Z minutes of boil. So I am guessing that I can change the power output to change the boiloff rate. There's going to be an upper limit where the kettle would boil over, and a lower limit where it stops boiling, but I'm guessing there is a range where I can control BTUs to the kettle to get a variable boiloff rate, and use PID control to steer it towards the target value.

I'm not sure you are going to get very accurate volume measurements based off of level sensing a rolling boil. I think you would need to take a bunch of readings over a few seconds or more and hopefully the average if all those readings is close to the actual level if it wasn't a rolling boil. The level in my sight glass bounces around by +/- 1/4 gallon when I'm at a gentle rolling boil. If I turn my heating element on 100%, then it's more like +/- 1 gallon. I think you will find that the surface moves around too much during the boil to accurately measure the volume.

Now with that said I think you will also have to wait a significant amount of time to measure a large enough delta in volume to get an accurate calculation of the current boil off rate. If your measurement accuracy is only say good to 1/8 gallon and you are trying to establish proper boil off rate within the first 15 min of boil then you are trying to measure a change that is only twice your measurement accuracy. The signal to noise ratio is going to be way too low. If on the other hand you wait longer like 30 min into the boil to calculate the boil off rate then you may be too late to correct it if it's off by a lot.

Consistent boil off rates is one of the reasons I love my electric setup. I just set it to 70% duty cycle and I get 1 gallon per hour every time. No need for any feedback control.

 

[rpierce]

Regarding using smaller pipes: what pressure are you using? I need to 0.4 PSI / 11" W.C. I thought that for low pressure, to get the same BTU flow, you need bigger pipes than higher pressure propane. I'm using jet burners designed for this pressure, and they are equipped with 1/2" NPT, so I figure going smaller would limit them further.

 

[rpierce]

I'm definitely planning on averaging readings. The level sensing isn't the same as a sight glass. It is measuring the pressure in a tube of air that is displacing wort. So it is a function of the mass of the fluid, which I don't think is going to change during a rolling boil. That said, I haven't actually built this part yet so I have no idea what will happen.

Either way, I need a way of throttling the boil. If I'm using a needle valve hooked to a stepper motor, that might be repeatable too. I won't know until I try it.

 

[tob77]

I used an RC servo motor to control the valve on my propane regulator with a raspberry pi. The main purpose of my servo is to control the mash temp. I have a setting in my program that sets the position of the valve at a fixed point during the boil. However, I have found that my boil off rate does change a bid from brew to brew depending on the weather. My propane regulator's valve is not very linear, especially at it's min and max positions. I wen't through a lengthy process of linearizing the valve by measuring the rate of change of the temperature at various valve positions. Check out my thread for more details:
https://www.homebrewtalk.com/f235/raspberry-pi-brew-controller-451059/

 

[BigBlock]

I'll throw out another option that I'm planning on trying. I'm using 2 parralel solenoid valves with an orifice (small in one and larger in the other)in each line to run a low/med/high heat setup. Once you get to a boil you can run on a the low setting and cycle the other valve to maintain the boil since the system has a pretty long time constant.
My main concern is how to determine when you have a good rolling boil.my plan to maintain the boil was to use a bubbler level sensor and measure the amount of noise. My assumption is that during the boil there'll be an increasing amount of noise on the sensor reading. if anyone has tried this I'd be interested in hearing their results.
in terms of accurately measuring the amount of water that's being boiled off I would think that measuring weight of your boil pot would be the straight forward answer as it will not be affected by boiling.

 

[craigmw]

Seems to me you could use Tob's method to control gas flow to the burner. Better yet, you could use a Maxitrol modulating gas valve, as is used in variable output furnaces:

http://www.supplyhouse.com/Maxitrol-M511-1-2-1-2-Modulating-Gas-Valve

For these, I believe you still need a solenoid valve upstream (or downstream, depending on the model) in the gas flow to completely shut off gas flow. The modulating valve controls the output depending on the voltage sent to it.

 

[rpierce]

The Maxitrol modulating gas valve sounds like exactly what I need, but I'd really not like to spend that much per valve! At this point, I plan to try the Parker instrumentation valve I mentioned and mate a stepper to it. I've already purchased the valves. I seem to recall that they have either 4 or 8 full revolutions of travel, so even if only the first fraction of the valve causes a difference in heat output, that's still a lot of room for a stepper to operate.

I'm also very interested in whether anyone else has used a bubbler style level sensor on a boil kettle and what results they got. Also, do bubblers introduce any significant hot side aeration?