Non contact flow sensor for heavy oil ideas please!!

Hi,

I am looking to investigate ways of creating a non contact flow sensor for a diesel fuel pipe. I have seen some on the market but they are real expensive, but are based around an ultrasonic sensor. In my eyes these are pretty cheap so I assume the cost is in the methodologies behind processing the information.

I dont need terribly accurate flow level values, even sensing whether a liquid is in the pipe or not would be sufficient (for now) to see whether it is economically possible.

Other than ultrasonic is there another way??

thanks

For water flow I've seen it done by putting something in the pipe that was spun by the passing flow and then using a magnetic pickup to detect the motion. In it's simplest form this gives you an indication of speed but not direction. I suppose you could get direction too by adding an extra pickup.

Its got to really be non contact and non invasive, so in other words nothing can go in or in line with the pipe, and be a clamp on kind of system....Only thing that comes to mind is ultrasonic but I dont know how reliable its likely to be..

idea 1: Could be a glass pipe, LED on one side and an LDR on the other. If there is diesel in the pipe it will scatter the light differently from no diesel.

idea2: A flexible hose on a scale. If there is diesel in the hose the weight is higher.

idea 3: in pipe pressure sensor, with a needle; can be read optically.

Where does the diesel come from, where does it go to, what is a typical flow?

Non-contact Ultrasonic devices are operating on doppler shift. The moving fuel will reflect the signal back with some small frequency shift. Just the same as sonograms and echo-cardiography. If you can find an appropriate source and have the ability to determine the frequency shift, and can do the math for flow through that particular pipe... Yeah, it's doable.

LDV is possible and no it won't be cheap (but probably a lot cheaper than when I first heard about it 30 years ago) http://en.wikipedia.org/wiki/Laser_Doppler_velocimetry

[edit - if the oil is transparent enough...]

Ultrasonic transit time type sensors are easier to do than doppler type.

rx tx __^^^____vvv_ \ / --flow---> \ / ________\/____

The ultrasonic signal travels against the flow and the delay can be correlated to the flow rate. A better arrangement is to transmit signals both directions and subtract the transit delay to get a more accurate reading. It is conceivable you could attach the transducers from a standard ultrasonic rangefinder to the side of a pipe, and experimentally calibrate its reading to flow rates.

You might try a simple electret microphone or a piezo transducer to sense the vibrations of liquid flowing through the pipe.

If there is ANY form of obstruction in the pipe under consideration (valve, coupling, restriction etc) a flow of fluid will create a sound. On the basis that all you are looking for is a yes/no flow signal, all you need is to clamp a miniature microphone to the line and do a spectrum analysis (frequency versus flow) to determine a suitable range of "detectable" signal filters. You will need filters to ensure that the various background sounds are ignored.

Oops, previous entry was lodged as I was typing. As they say, great minds think alike :D

notorious: Its got to really be non contact and non invasive, so in other words nothing can go in or in line with the pipe, and be a clamp on kind of system....Only thing that comes to mind is ultrasonic but I dont know how reliable its likely to be..

Can you describe the measuring conditions in a bit more detail? What sort of pipe dimensions and fluid speed are you talking about measuring? Is it diesel fuel, or something else? How accurate do the measurements need to be? Exactly what are the restrictions on modifying the size/shape/routing of the hose to incorporate the sensor?

coriolis mass flow meter will work nicely. However you most likely can't afford one.

http://en.wikipedia.org/wiki/Mass_flow_meter#Operating_principle_of_a_coriolis_flow_meter

Used them in a refinery I worked in. Best, most accurate flow meter on the market, is not effected by the chemical composition (specific gravity, viscosity, temperature, etc) of the material being measured. They have come down in price in recent years but still pricey and beyond most hobbyist budgets.

Lefty

Thanks for the feedback guys.

Any in pipe option is not possible. Were talking a Diesel supply pipe exiting a tank which has a diameter or about 3 inches or so, and made of steel. The non contact flow meter is to be used as an alarm to prevent fuel theft. Logging how much has gone isnt required, just an indication to show fuel is moving in the pipe when it shouldnt be.

As its only got to be stationary or flowing the sensor needs to be cheap and cheerful. I like the microphone idea although removing background noise to prevent false alarm will be a challenge, as it will probably be the cheapest to implement. Ultrasonic seems to be the only other viable option within budget, unless there are other methods...

A couple of ways come to mind. When I did this for the home boiler, I simply measured the pressure that the fuel pump delivered and then used the Fuel flow chart from the nozzle manufacturer to determine the rate of oil flow. A simple current switch on the fuel pump then allowed fairly accurate open-loop oil consumption measurement.

Can you consider this sort of approach or is your equipment of the modulating kind? Current switches are cheap & noninvasive.

If fuel flow is variable, you can add a pressure sensor on the high side of the fuel pump to determine the pressure and from there extrapolate the fuel consumption.

notorious: Thanks for the feedback guys.

Any in pipe option is not possible. Were talking a Diesel supply pipe exiting a tank which has a diameter or about 3 inches or so, and made of steel. The non contact flow meter is to be used as an alarm to prevent fuel theft. Logging how much has gone isnt required, just an indication to show fuel is moving in the pipe when it shouldnt be.

As its only got to be stationary or flowing the sensor needs to be cheap and cheerful. I like the microphone idea although removing background noise to prevent false alarm will be a challenge, as it will probably be the cheapest to implement. Ultrasonic seems to be the only other viable option within budget, unless there are other methods...

What sort of minimum flow speed do you need to detect?

Its around a 3 inch pipe and if the detector can sense within a gallon flowing past I would be happy.

There are some heat-based flow detectors on Omega.com. Might be worth investigating. That said, with a 3" pipe, you may have some serious flow variation based on the viscosity of the fluid and the speed at which it is traveling. Friction at the pipe inner surface will slow the flow down significantly from that at the center of the pipe. You will have fun calibrating all that.

If it was me, I'd check and see if there wasn't another way to do it, i.e. put a speed sensor on the pump motor, etc.

How about measuring the level of diesel in the tank with an ultrasonic sensor?

When you say a gallon, do you mean a gallon a minute or a gallon a month? Is it actually displacement you want to measure rather than flow rate? If so, perhaps what you really want is the sort of fuel gauge they use in commercial fuel pumps.

If you’re trying to prevent theft then proximity sensors with alarms would probably be a better answer. A locking valve wouldn’t hurt either.

Shame about the non-invasive part.

Noninvasive microphone measurement of gas flow in residential installation described here: http://dub.washington.edu/projects/gassense