Problem of Radar Level Sensor in Cold Weather

[PW2006]

We installed several radar level sensors on the outdoor tanks. They are all identical. The installation location on the tank roof is correct. They worked fine in the summer and autumn. But when the weather gets very cold below 30oF, mostly in the night, they start giving false reading frequently here and there (I can not say that it is everytime and on every sensor). But when the temperature becomes warmer (over 32oF mostly in the day time), they automatically go back to normal and correct readings. It is weird. We have heavy oil in the tanks. For example, if the sensor reads the tank as 90% full during the day time. In the night, when the temperature drops down below 30oF, it reads the same tank level as 100% full. When the temperature goes up again in the day time of next day, the reading on the same sensor goes back to 90% full automatically.

Per the sensor spec., it should work at even much cold weather. The sensor manufacturer has its own guess.

Do you have any experience with this problem? Do you have an explanation and a fix of this problem?

Thanks.

 

[itsmoked]

I'd guess that you are having water freeze across the face maybe. Especially since it happens pretty much at the freezing point. Can you run any kind of heater near it? Even a light bulb? Something that will always keep it above the dew point.

Anything just to do a test.

I realize this is next to oil so you may have issues with ignition sources.

Maybe an explosion proof bar heater. Something the raise the temp about 10F right near the radar head.



Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[PW2006]

Our oil won't freeze and there is no water in the tank. Tank is vented to the air. The sensor points downward. Do you mean that moisture is condensed from the air at first and it freeze on the surface of the sensor head?

 

[itsmoked]

Yes! That's my logic.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[itsmoked]

There is probably some sort of radar clear cover over the electronics that is getting a frost over it, or condensate that freezes and the unit just reads its own face distance. Could even have water sitting on the top of its face inside the unit.

It may see thru it until it freezes. Then turns clear again when solidly frozen and sees thru it again??? Heating its case slightly will prevent condensation that could be occurring on the outside OR inside.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[MacGyverS2000]

Plenty of moisture in the air, even if it's not mixing with the oil itself... a little dab of dew will do ya.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

That scenario is not at all unreasonable. During the day, the tank heats up and air vents out. Given the thermal lag, by the time the tank cools down enough to suck air in, the inlet may be swamped with dew already, causing massive amounts of moisture to enter the tank. After a couple days of this you could actually get quite a bit of liquid water inside the tank. Additionally, due to the radiated thermal behavior, it's possible that the top of the tank (and the sensor) can get below freezing even when the air temps stay above freezing.

We had a system that would routinely collect cups of water after several days of humid days and cool nights.

TTFN

FAQ731-376

 

[Comcokid]

I know from working with the TDR type level sensors that moisture will dew-out on the probe even in vented fuel or diesel tanks with no measurable condensed water on the bottom. Even though the tank is vented, it can still have a higher relative humidity in the tank. Add a temperature drop, and the humidity will exceed 100% relative. Add enough temperature drop and the dew that forms will freeze.

In TDR sensors, the false-reading effect is similar if there is enough condensation. One thing you can try is to tap the sensor, or the tank near the sensor when it is giving false readings. If it is dew, the drops will be super-large and a rap will cause them to drop off.

 

[OperaHouse]

I bet this is an insulated tank and the location of the sensor is not. This will cause the sensor to be the first place condensation forms. Even if it isn't, insulate around the sensor area for a couple feet to it is the last place to cool. Then the rest of the tank will condense the moisture first. A can of foam could work wonders along with a styrofoam rose cover.

 

[danw2]

For what it's worth . . .

We've seen false high readings due to condensed water, too.
I sat in on a meeting with the Siemens sales guy today.

He showed a couple slides of their test stand with a Siemens LR250 with the horn antenna and screen shots of echo profiles for both dry and wet conditions. The photo showed a hand pumping a sprayer to coat the horn with water to create a 'wet' conditoin.

The dynamic baseline shifted between dry and wet conditions, but the peak level reading did not.

Apparently Siemens has done some signal processing that avoids the problem of false high readings due to moisture on the antenna.

Apparently, the capabilities reside with the LR250 with the 2" mount. The 1.5" LR250 is different.

 

[itsmoked]

Hello? Hello? click click click Hello? Hello?
Is there anybody there??

I think the line is dead...

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[PW2006]

Sounds interesting, danw2.

Our radar sensors are made by E+H. They told me that they could adjust the settings of the sensor to solve this problem. But it has to be done with their technician and when the sensor is having problem in the cold weather. It is not free. I have not asked them to do it yet.

We considered using Siemens sensor before. Your information is useful. But I guess our ice condition is still different from the wet condition.

Thank all you guys, danw2, itsmoked, Comcokid, Operahouse, IRstuff, and macgyvers2000.