Eastman Chemical's main facility in Kingsport, Tenn., is a huge challenge for anyone responsible for monitoring flow in its miles of pipe. Started in 1920, the complex now stretches over 4,000 acres and contains 550 buildings. The main plant site alone covers 900 acres. The facility makes a wide range of products to support Eastman's five business segments. So, not surprisingly, it boasts every type of piping imaginable, carrying gases, slurries and an incredible number of industrial chemicals.
Because we have such a variety of piping and products, we periodically reassess our measurement capabilities. In 2009, we decided to thoroughly investigate flow metering to see if we could benefit from recent advances.
Much of my job as an environmental and process analytics chemist is to use portable devices to ensure accuracy of installed flow meters, to troubleshoot process upsets and to do flow checks on unmetered lines. Accuracy is my main concern. So, I try my best to keep up with the latest technology.
The need for accurate, representative flow data that we could archive and access has been increasing exponentially, but nothing in place could provide all the information we required. Several internal development and engineering groups had looked independently into earlier-generation clamp-on ultrasonic flow meters with data logging capability and had varying degrees of success with them. For the most part these meters were collecting dust on a shelf. The problem was accuracy. I often heard things such as: "We've had a clamp-on ultrasonic for years and I've never had much luck with the thing" or "Clamp-on meters are very frustrating to use. You never know if they are giving you accurate readings, if you actually get any readings at all." I believed such meters had vast potential but hadn't lived up to it. So, I resolved to see if there had been any advances.
I called every clamp-on ultrasonic flow meter manufacturer I could identify. My intent was to get these meters in-house and test each of them on process and utilities pipes. I selected 20 points throughout the site; several are in our coal gasification plant, which transports gases and liquids as well as slurries of changing consistency and temperature.
HOW ULTRASONIC WORKS
The technique most ultrasonic flow meters use is called transit-time difference. The meter sends ultrasonic pulses through the medium, one in the flow direction and one against it. Transducers alternate as emitters and receivers. The transit time of the signal going with the flow is shorter than the one going against. The meter measures transit-time difference and determines the average flow velocity of the medium. Because ultrasonic signals propagate in solids, you can mount the meter directly on a pipe and measure flow non-invasively, eliminating any need to cut the pipe.
The tests. One of my first tests for all the meters was in a parking lot where we have a 30-in water line. Aside from accuracy, I focused on ease of installation. After all, I have to go all over the plant measuring flow and I don't want to waste time setting up. I also was looking for reliability, data logging capability, diagnostic tools and good battery life.
The water line provided telling results. If a meter took too long to set up or didn't give adequate accuracy on such an easy task, there was no point trying it at the other test sites. Some meters didn't make it out of the parking lot.
And the winner is… I'm not going to name those meters or, for that matter, the others I didn't ultimately select -- some were good products with good accuracy and I don't want to make them look bad. But, at the end of six months, one meter met all of my standards, especially for ease of set up and accuracy. It was from Flexim.
It had two distinct advantages over the others. First, it is a combination gas/fluid meter; so I only had to buy one meter, not two. Second, its installation is a snap. The main problem with the others was installation. Flexim has a fixed frequency on its transducers. Regardless of what's being measured, they keep that frequency. The others use a sweep frequency -- it takes quite a bit of time to set up their transducers and put them on the pipe. Then you have to find a "sweet spot" and, even after you've found the spot, you still have to adjust the transducers to get an accurate measurement.
There are many ways of getting a signal to a pipe. Lots of meters rely on multiple transducers that operate at different frequencies to cope with the various types and wall thicknesses of pipes and diverse liquids. Flexim uses the same transducer for everything but filters the transmission pattern and either increases the transducer voltage (which can range from 15 V to 90 V), or breaks up the pattern of transmissions to adapt to real-world conditions. The meter sends approximately 1,000 pulses back per second in 500 pairs; the system automatically recognizes a change in liquid consistency. Not even coal slurries are a big challenge. The system, through a combination of software, clever transducers and signal processing, automatically realizes it's losing signal and boosts power to the transducers. It can monitor anything from ¼-in tubing to a 30-ft penstock.
I think it's safe to say that ultrasonic flow meters finally have achieved their potential. I have taken more than 600 readings -- including on pipes containing acids, water, gases, and slurries at several hundred degrees -- with this portable meter and my success rate is 99.8%. (The lead photo shows me checking a 42-in-dia. line.) At the end of the day, I collect all the data dumps from the built-in data logger and export them to a spreadsheet.
Our internal customers have gained faith in ultrasonics and have bought ultrasonic units to replace their older meters. Concerns about accuracy are a thing of the past. One of our clients was running an addition process and having some problems with the feed rates. An orifice plate flow meter on the line wasn't giving consistent measurements. I took the portable over there and spent three days. The client liked the results but wasn't sure about the accuracy. So, I brought the Flexim meter to our metrology group, which has a Coriolis flow meter set-up, and we bench tested the portable against it. The correlation was 0.9999.
We currently have five ultrasonic meters in our group -- four Fluxus F601 liquid units and one Fluxus G601 gas unit (Figure 1), which actually can measure liquids as well. Most applications we have are for liquid measurements but with a plant this size the ability to measure gas flows from outside the pipe is definitely a plus. The G601's specifications give a 100-psi minimum requirement to measure gas flows. The lowest-pressure line I've personally attempted to measure is 140 psi; that particular application was for a nitrogen header and the meter worked well.
At the present time our Flow Measurements Group consists of three people (two utility operators and me). All across the site, people are clamoring for us to come over with the clamp-on meters. This has prompted us to do some "creative scheduling," but that's a good problem to have. Ultimately, we may add people to the group.
We continually are finding new applications for this technology. When we first started using the devices we were doing straightforward chores -- basically chemical-addition-type measurements expressed as gal/min or lb/min. We also did a fair amount of checking and troubleshooting existing flow meters. We have begun to expand our repertoire by adding heat flow (BTU) type measurements. The meters are dual channel and configured for temperature inputs, allowing us to measure supplies and returns simultaneously. We never had this energy meter functionality before and it's already starting to pay off. It's definitely something we'll keep our eyes on.
Where I used to hear "Ultrasonics might as well be Ouija boards," now all I hear is "What's my flow?" Ultrasonic is definitely the way to go today.
GREG HARPER is an environmental and process analytics chemist and leads the Flow Measurement Group at Eastman Chemical's facility in Kingsport, Tenn. E-mail him at email@example.com.