Large LP Gas Tank Fuel Monitor via Transducer

[AnthonyZ]

So I've been doing more research on this since this thread came up (oh how I wish it were in full swing BEFORE I spoke to the propane service provider for my install...). It seems to be a relatively simple (and not at all unusual) application. The vast majority of the commercial products available are using hall effect sensors at the gauge. In other words, most use a float level meter and sense the hall effect, parse it and send the signal in various forms from there. I am going to return to my project next week with a handful of linear hall effect sensors and, to be prepared, a handful of pots. I'll then crack a gauge and see what works best to measure the level.

 

[BraveSirRobbin]

I can't believe there isn't an off the shelf product for this application! What manual gauges are you thinking of using? Are you taping into the 'dial' mechanism? If so, how do you know it will provide enough torque to turn a linear pot?

 

[AnthonyZ]

There are a bunch of commercially available products for this type of project. Most seem to be designed to be installed prior to deployment but, there are retrofit options. Having studied 'em, I'm gonna roll my own. I found a ton of stuff yesterday but I saved 'em on my kid's machine. I'll pop back in later with some useful links. As to the torque, I don't know. That's why I'm gonna return prepared with a bunch of various hall effect sensors (I bought 5 different models in groups of four from Digikey). Ultimately, as I stated earlier, the project I built seems to be tracking well via pressure meters currently but, if there is any reason to doubt it (and there is), I'm gonna see it to the end if it kills me.

 

[felixrosbergen]

So on my in ground LP tank there's a guage that tell me the % Full. Is this then measured by a float connected to the guage rather then pressure? Guess it woudl make sense.

Measuring the tank is one thing, but hwo anybody measured the flow?

I woud like to get a 'total energy input' picture of my house. For the electrical the new brultech unit seems more than sufficeint, but on the gas i don't know what to do.

 

[sda]

So on my in ground LP tank there's a guage that tell me the % Full. Is this then measured by a float connected to the guage rather then pressure? Guess it woudl make sense.

Measuring the tank is one thing, but hwo anybody measured the flow?

I woud like to get a 'total energy input' picture of my house. For the electrical the new brultech unit seems more than sufficeint, but on the gas i don't know what to do.

pulse count gas meter. most are rated for both nat gas and propane.

 

[sda]

some ideas:

http://www.squibbtaylor.com/subfiles/taylorcatalog.pdf

 

[SteveQ]

BSR

I'm not sure what kind of accuracy you need, but you can also calculate the amount of propane you are using based on the time your furnace is ON. I have been doing this for the past 2 years and I believe the the data is quite good. I have a propane furnace that generates 77,000 BTUs/hr ( This info is on the rating label for the furnace) . Propane produces 91,500 BTUs/gal. Therefore my furnace uses 0.842gal/hr. I use Homeseer and a x10 powerflash module to measure the ON time of the furnace. I keep a running total. When the furnace has run for 500 hours, I know I have used about 410 gallons of my 500 gal tank.

Maybe this approach could work for your application.

Steve Q

Note: I also have a gas cooktop, but the amount of gas it uses is so small it does not make any significant difference.

 

[felixrosbergen]

my hot water heater is LP too, but it has a exhaust fan on it that uses electricity so i could monitor that to determine if it's on.

My gas service company just keeps the tank full but stopping by every x amount of time to top of the tank. In the winter this is every 2 weeks or so and in the summer i imagine much less.

The proble is that i don't necesarilly know when they come and top off. So while your method is good to determine the house energy consumption it won't tell me how full my tank is. With the service i don't really have to worry about it...but still nice to know in case they screw up.

 

[WayneW]

The proble is that i don't necesarilly know when they come and top off.

My LP tank is buried, so I could put a mercury switch in the tank lid with a DS10a or a security transmitter. Might you have a way to know when they access your fill connection area? Maybe build a monitored "weather" cover or a photobeam?

 

[SteveQ]

Wayne,

I like the idea of a mercury switch/DS10 on the tank lid. This would tell me if they came and filled the tank. I would use the DS10 to trigger an event to send me an email - "Tank Filled". I would also have it tell me how long the lid was open. This would allow me to know if they just checked the level versus actually filling the tank.

I would never have thought of this!!!

Steve Q

 

[DavidL]

I know when the propane company comes....they leave with a big fat bill behind hanging on the door knob

 

[AnthonyZ]

I know when the propane company comes....they leave with a big fat bill behind hanging on the door knob

Now that's funny!

 

[AnthonyZ]

Alright, so I've finally been able to return to my metering project and I came armed with some hall effect sensors from Allegro. I've got some simple code set up in a micro controller to read the analog input from a linear hall effect sensor that I then "print" every 100ms. The tanks have analog meters on them that, conveniently enough, use magnetics on the float meter in the tank to spin the dial on the circular, analog meter. I pulled a meter out and found that by placing steel (a screwdriver) in the vicinity, I can spin the dial through it's whole range. Cool. I then set up a dummy rig to use the micro controller to measure what it's seeing off the meter's needle. The hall effect sensor can be placed anywhere I want above the meter. The issue I now have from a coding point of view is that as the magnet in the meter spins, the empty and full positions are very near one another. This means that I get similar readings from the hall effect sensor when the tank is near full or near empty. I'm, trying to think of some logic to compensate but, I haven't wrapped my mind around it. I think I may use two sensors above each tank to compensate. The good news is that hall effect can easily be used (once I figure out how to code it) to get a very accurate measurement. Any ideas on how to deal with the "circular" logic are welcome.

 

[znelbok]

Having just completed a similar project last week, I can chime in. I had a client with a remotely located cabin in the mountains of CO who wanted to be able to monitor his three separate propane tanks from TX. The first thing I did was contact the propane utility company that services his home and start asking questions about how best to go about measuring the level. Their suggestion was to use pressure and to obviously measure prior to the regulators on the tanks. I ordered some Wika pressure transducers from Davis and sprung for the intrinsically safe model as I did NOT want to blow anything up. These were installed at the tanks (all three are arranged about 6 feet from one another). They were 4-20mA models and thus, I just wired two wires from each meter back to an enclosure I mounted to a 4"x4" set with concrete into the ground. The outputs of the meters fed analog inputs on an Arduino micro controller with an Xbee shield and the whole shebang was powered off of solar panels purchased from Sparkfun Electronics. I also built a battery charging circuit feeding a 12v7ah "security" battery so that I could get a couple of days of monitoring without sun. The Xbee modules are using the Zigbee AT command set and the send unit is transmitting serial data from the Arduino about 200' to the cabin. There, the Xbee is connected to a MaxSerial Arduino board from Fundamental Logic and it is connected to the client's HAI panel. This allows for an update of the tank levels every 5 seconds and, using messages I update the HAI and Convergent Living systems. I also use WebLink on a local server (that's up 24/7 to also handle a number of Panny IP camera's DVR functionality). WebLink (and the client's SnapLinks) allow him to monitor messages (indicating levels) and also sends emails when the tanks are critically low. Since the cabin is so remote, we consider anything below 25% as critical and the emails come every additional 5% drop after that. So far, it's only been up a week but, thus far, it's working well. Also, the pressure meters are tracking linearly (fingers crossed, I may need to add some temperature offsets to the Arduino's C++'ish code) with the tank's on board analog meters.

A cooment on this so others dont do it wrong as well.

Instrinsically safe (I.S.) devices such as the transmiiter used here need to be designed as a system. AN I.S. transmitter needs to be used in conjunction with a barrier. A barrier is an energy limiting device that is also certified.

The barrier has to be matched to the transmitter and cable characteristics have to be taken into account as well

There are also two levels of I.S., Exia & Exib

The zoning of the hazadous area (Zone 0, 1 or 2) that the device is being placed in needs to be determined an that dictates what transmitter I.S> level can be used.

There are also temperature considerations, type of Gass and power levels.

In Summary, the above installation is NOT SAFE AT ALL.

Please use a compentant person for hazardous area installs. Sometimes the gas companies will have their own regulations that mitigate the need to use I.S., but they have done similar work to ensure that the system is safe.

Mick

 

[AnthonyZ]

In Summary, the above installation is NOT SAFE AT ALL.

[/quote]
Actually, Mick, you're dead wrong. You didn't bother to ask what was in place at the tanks, what level of intrinsically safe meter was used or how it was cabled before you posted your opinion in such strong terms. The meters used were sealed, intrinsically safe models designed for hazardous gas installations and all safety precautions were followed to a T. Thanks for your input, though.