compensate pressure sensor for temperature differences

[Toolebox]

I am currently working on a project for my home brewery using a V350-35-ta24.

In my boil kettle (15 gallons), I am using a fairly cheap direct screw in 4-20ma pressure sensor to calculate water volume using linearization to convert the value to gallons.  This works well, until i start to heat the boil kettle.  The pressure sensor value increases based on temperature increase, which will throw off other calculations i intend to implement down the road.

I have attached my project.  If you look at the first net in the main routine I have compensated for this by adding a second linearization based on the difference in gallons between cold starting  water temp and 200 degrees F, then subtracting this number from the original value to find actual volume.  This is obviously not perfect, as i assume different volumes of water will have slightly different results, but was the best i could come up with to get things close.  The volume is also used for safety checks, to make sure that the heating element contactor turns off when there is not enough water.  So its important that its as accurate as possible

I am new to PLC programming.  when you look at my project you may find things are not done perfectly, but everything is labeled well and currently working other than the volume is less accurate than i would like.

What is the best way to compensate for the temperature change to keep the volume measurement consistent?

Any advice would be greatly appreciated!

V350 Brewery.vlp

[Joe Tauser]

8 hours ago, Toolebox said:

I am using a fairly cheap direct screw in 4-20ma pressure sensor to calculate water volume

For reference, post the make an model of the sensor so we can see it's specs.  

This isn't something you can program around.  You need to modify the installation of the sensor.

Isolate the sensor from the high temperature fluid.  You're most probably raising the temperature of the electronics inside the sensor to a place they were never designed to operate and it's screaming at you.  I am a distributor for Dwyer and the low end sensors I typically use top out at 175F.   You will eventually destroy the amplifier in the sensor if you keep taking it above 200F.

This is how the big boys do it at breweries, oil refineries, and other places with nasty extremes:

https://www.emerson.com/documents/automation/product-data-sheet-rosemount-dp-level-transmitters-1199-seal-systems-en-73226.pdf

Now if I was rolling my own brewery and I was on a limited budget and knowing what I know, I'd make a simple 1/4" extension pipe a foot or so long to move the sensor away from the tank and the heat.  I'd keep the pipe fairly level to the opening so as not to screw up the reading.  Then while I was increasing the temperature I'd figure out a way to keep the fluid in that pipe  relatively cool, like attaching a heat sink of some kind or even throwing a wet rag over it with a fan on it.

Joe T.   

[Toolebox]

Joe,

Thanks!  Here is a link to the sensor I am using http://www.sendo-sensor.com/pro1/19_100.html.  the specs show it is rated for the temperature, but maybe it cant handle heat variance well?  I do have a Dwyer 673-2 pressure sensor kicking around I could hook up via silicone tubing for an air gap, but figured that the one mentioned above mounted directly would be much more sanitary / easy to clean.

Do you believe that I would have the same issue with the Dwyer sensor raised above the kettle, but connected to a fitting at the bottom of the kettle via silicone tubing?  I could always use that until I settle on a better sensor that can handle temperature differences better?  Otherwise, I will attempt to extend the Sendo sensor out via a copper pipe and come up with some sort of cooling.

Really appreciate your help.

-Ryan

[Joe Tauser]

I'd stick with the sensor you have as it has a nice, flat, easy-to-clean diaphragm.

The specs show it is rated for 100 C media, or the temperature of the fluid.  They also say the maximum compensated ambient is 70 C.  This is the maximum the housing and subsequent electronics can withstand.  Your process connection is probably pumping heat into the housing faster than it can be dissipated.

7 hours ago, Toolebox said:

sensor raised above the kettle, but connected to a fitting at the bottom of the kettle via silicone tubing

Uh, no.  Think about this configuration.  You're measuring head pressure, and sensing element must be at the level of the pressure you want to read.  If you go up, the weight of the fluid in the tube will counteract the head pressure of the tank and null it out.  And you can't use air as a transfer medium because it's compressible.

Have you ever siphoned water out of a fish tank?  If you put a tube in the tank, water doesn't come out of the end while it's above the water.  Water doesn't come out if you lower the end below the water level, either, because it's full of air which is "springy".  It's only after you suck on the tube to fill it with water AND put the tube below the water level can you take advantage of the hydrostatic pressure (head) to get the water flowing.

The pressure sensor MUST be at the same level of the port.  You can use a long length of silicon tubing to get the heat out, but you'll need to put a tee and a valve at the end where the sensor is so you can purge the air out.

7 hours ago, Toolebox said:

mounted directly would be much more sanitary / easy to clean.

This is very true and if you were making beer to sell and had to worry about the FDA inspector then you would need to install a sanitary fitting.  The diaphragm in the link I posted is filled with silicon fluid and sealed when attached to the sensor, which lets you get away with moving the sensor around as there's no way the volume can change in the pressure transfer conduit.

So you are going to have to take this thing apart and clean / sanitize it from time to time.  So keep your tubing as short as you can so you can run a round brush through it and then throw everything (but the sensor) in the dishwasher to get it hot enough to kill the bugs.

If all this is a problem I'd look into another technology for level sensing.  The only problem is you can spend a lot of money in a short amount of time on harsh-duty food grade sensors.  Here in St. Louis, there is a reason companies that sell that sort of equipment have dedicated salespeople for Anheuser Busch.

Joe T.

[Toolebox]

Thanks for the additional info!  ill use the sensor i have and add a pipe and cool the pipe if needed.

 

Thanks again for your help!

-Ryan

[Ausman]

You could always go outside the square and use a cheap load cell and transducer to measure the weight of your pre-intoxicater.  Of course you have to allow for all the other associated bits hanging off/attached to the kettle, but the end result should be a completely accurate volume level.  No cleaning or other hassles.  No point having a nice home brew if it kills you.

Not being a typical Aussie I don't know how much beer changes weight as it brews.  I do know that too much beer increases weight somewhere else!  🍺  

I think I developed a dislike for the stuff after having to go into pubs first thing the next morning to fix things, and reeling at the pong.....(smell for you non-aussies).

cheers, Aus