1 December 2005

Coriolis technology is elegant, tough

New signal-processing techniques conquer two-phase flow barrier.

By Wade Mattar

Coriolis technology offers unprecedented accuracy and reliability in measuring material flow.

It may be the most superior flow measurement technology. However, conventional Coriolis meters have had one significant limitation: They haven't performed well in measuring two-phase flow conditions, flow that involves a combination of gas and liquid mass.

Two-phase flow can cause process interruptions and measurement inaccuracies that can significantly affect production and profitability. Recent developments in digital Coriolis technology overcome the challenges of measuring two-phase flow to improve traditional pipeline flow measurement, while offering a solution for demanding applications that have been traditionally out of reach.

Partially filled flow tube

With worldwide revenues presently greater than $400 million and moving to $600 million in the near future, Coriolis meters are among the fastest-growing flow measurement technologies. These meters measure flow by analyzing changes in Coriolis force of a flowing substance.

Coriolis force arises from a mass moving within a rotating frame of reference. That rotation produces an angular, outward acceleration, together with linear velocity, to define the Coriolis force. With a fluid mass, the Coriolis force is proportional to the mass flow rate of that fluid.

To use Coriolis force for measurement, a Coriolis meter has two main components: an oscillating flow tube equipped with sensors and drivers, and an electronic transmitter that controls the oscillations, analyzes the results, and transmits the information.

Reliable Coriolis measurement depends on consistent, reliable oscillation, which four factors determine: the density of the liquid, the balance of the tubes, the dampening caused by the flow stream itself, and the physical isolation of the tubes from the environment. Compromising even one of these factors will degrade Coriolis meter performance. Yet two-phase flow compromises every one of them. Thus applications involving negligible amounts of entrained gas-even as little as 2% volume-have been poor candidates for Coriolis measurement. This has been particularly troubling in applications where reliable, highly accurate flow measurement can confer considerable bottom-line advantage, but where two-phase flow is an integral part of the process or it is necessary to begin with an empty or partially filled flow tube.

Making matters worse, entrained air may not emerge as the culprit until after a frazzled process engineer has invested many hours trying to figure out why he can't get the results he needs. Our own analysis shows up to 92% of all Coriolis measurement problems are due to entrained air or gas, yet in the vast majority of cases, users don't recognize two-phase flow as the problem.

Patented and advanced control

Coriolis technology is highly accurate in single-phase flow, with perhaps a ± 0.1% error level. Two-phase flow can boost the error rate to 20% or higher. Following are some of the profitability drains that inaccurate flow measurement causes:

Lost production: In flow-intensive operations, thousands of dollars' worth of lost production can pass undetected in minutes.

Inaccurate pricing: In custody transfer applications, where measured amount transferred defines payment price, faulty measurements raise financial havoc on either end of the transfer.

Excess downtime: When traditional Coriolis meters encounter entrained air, they render inaccurate measurements, and if the condition persists, will shut down. This cuts into valuable production time.

To determine the extent of the problem and find a cost-effective solution, Invensys commissioned a survey of process engineers. This revealed entrained air was indeed a major problem in the industry, and what customers really wanted was a cost-effective meter that provides accurate measurement despite the presence of air. Invensys collaborated with researchers at Oxford University in England to develop digital technology for accurate measurement of flow, even when the flow tube contains entrained air.

Working closely with the Oxford researchers, our engineers developed a transmitter that applied the Oxford measurement principles. The resulting patented product incorporates new signal processing techniques to provide useful measurements of both mass flow and density and the operational aspects of keeping the Coriolis meter running stably in single-phase or two-phase flow conditions.

One of the many patents it has received involves an advanced control and measurement system with high-speed digital signal processing that responds to changing flow conditions many times faster than standard Coriolis flowmeters. Another patent relates to detecting and compensating for two-phase flow conditions and generating a validated mass flow measurement.

Unloading railcars and trucks

Coriolis meters measure the mass flow of materials, which is independent of other physical parameters, as well as the ambient conditions in which the measurement takes place. Therefore, the measurement is unaffected by changes in temperature, pressure, density, viscosity, and flow profile. With the ability to handle two-phase flow and compensate for physical conditions, the advanced Coriolis flowmeters have greatly expanded fluid metering applications, including traditionally difficult situations, such as custody transfer, proving, tank truck, and tanker loading and unloading, and applications where two-phase flow is an integral part of the process.

Accurate measurement at custody transfer points is critical as competitive market conditions drive companies to develop operations that are more efficient. By minimizing wasted materials left on the bottom of transport vessels and improving transfer yields, advanced Coriolis flowmeters provide more accurate material accountability, which is a direct contribution to bottom line performance. This is a win-win situation for both entities involved in the transaction. Advanced Coriolis technology is increasingly replacing positive displacement meters for custody transfer to attain the benefits of Coriolis accuracy, while reducing total cost of ownership. With no moving parts in the fluid stream, Coriolis meters require little-to-no maintenance and are easy to install.

In pipeline flow measurement-proving applications, the frequency and duration of calibration can hinder productivity. Advanced digital Coriolis flowmeters offer a solution by providing a much faster response time, and greater accuracy and repeatable proving with small volume provers. A proving run may take as little as 20 seconds. This is particularly beneficial in multi-product pipeline applications where fluids varying from light, liquefied petroleum gases to heavy crude oils pass through a common flowmeter. For these applications, flowmeters often take place several times a day, so slashing each proving process to seconds can significantly boost productivity.

Another issue is unloading railcars and tank trucks until they are practically dry. To empty out the tank completely, invariably introduces air as the level approaches bottom. Exacerbating this is that in most cases, unloading happens at as high a flow rate as possible to speed up the process. This high flow rate tends to suck air into the flowmeter. Where a conventional Coriolis meter would shut down in this situation, advanced Coriolis meters continue to provide a useful flow measurement, enabling faster, more complete unloading of tank trucks and railcars. Even with the flow tube empty, they respond 10 times faster than traditional Coriolis transmitters, which reduce startup time while increasing production throughput and profitability.

Innovative moneymakers

In addition to improving existing flow measurement applications, advanced Coriolis technology is opening new doors for improving process efficiencies where two-phase flow is an integral part of the process.

For instance, using carbon dioxide (CO2) for enhanced oil recovery (EOR) can increase output by as much as 12%. However, accurate measurement of CO2 has been the weakness of the process. A large midstream energy company found the solution by applying advanced Coriolis metering technology as part of a three-stage EOR program. The first stage was primary oil recovery, based on natural gas driving the oil to wellheads. Secondary efforts involved water flood driven production using natural aquifers. As primary and secondary production methods declined in effectiveness, tertiary oil recovery techniques gained attention. A number of EOR options received scrutiny, and CO2 injection into the oil reservoirs was the most effective method for extracting and moving oil to the well bore. While the yields from this EOR were significant from the start, engineers felt they could do even better if they could more accurately measure the CO2 flows in each well.

The problem is when CO2 is above the critical point, it exists as a gas and is easily measurable with standard gas measuring devices such as orifice plates. However, below the critical point, it can coexist in two phases, liquid and gas.

The company transfers CO2 in pipelines to multiple injection wells throughout the field, and variations in ambient temperature and pressure outside the pipeline have a dramatic affect. On a cool morning, they could have primarily liquid CO2 in the pressurized distribution pipelines. However, in the afternoon, with elevated outside temperatures, they could have primarily gas.

Possible options considered were orifice plates with multivariable DP transmitters, Vortex meters, and conventional Coriolis flowmeters. While traditional Coriolis technology is highly accurate in single-phase flow, with a 0.1% plus or minus error level, two-phase flow can boost the error rate to 20% or higher. None of these options met the company's performance standard, so they explored new avenues of flow measurement technology.

The company tested an advanced digital Coriolis flowmeter to successfully measure two-phase CO2 and based on the results installed the flowmeters at each of the injection wells. The advanced Coriolis flowmeters improved the accuracy of CO2 measurement by 300%. This provided the immediate benefits of increasing oil output, as well as the long-term advantages of accurate flow measurement data to correlate optimum production efficiency with the volume of CO2 injected, which is critical for developing oil reservoir strategies.

The above cases are but a small sampling of the many ways in which the benefits of Coriolis accuracy can work areas that have been traditionally out of reach. Every day we are seeing new applications wherein advanced Coriolis flowmeters are working successfully to solve traditional problems.

So look at your flow measurement challenges. Advanced digital Coriolis technology may be the solution to today's problems and tomorrow's innovations.

Behind the byline

Wade Mattar (wade.mattar@ips.invensys.com) has degrees in fluid engineering and aeronautical engineering. He works in the area of flow measurement with Foxboro Company/ Invensys and has for more than 30 years. He holds several flowmeter related patents, has authored papers and handbook chapters, and he chairs or sits on a number of national, international and industry related flow standardization committees.

Details on the Coriolis phenomenon

Coriolis flow measurement is a comparatively new technology. The first commercial meters appeared in the 1970s. They measure mass flow directly with high accuracy and rangeability.

Read this tutorial on Coriolis technology: www.isa.org/Fundamentals/Coriolis.

Coriolis in lethal service

By William Barnett and Rusty Liner

Minimizing exposure to cyanide, phosgene by reducing the number of leak points is a plus.

For a small percentage of applications such as HC, Sodium Cyanide, Sulfur Dioxide, and Phosgene, there was no safe method when using this technology.

For decades, the industry's standard for flow measurement has been differential pressure (dP) flow.

dP consists of a head producing device (orifice plate, Venturi, flow nozzle, Pitot tube), a differential pressure sensing transmitter, and the associated tubing/connections.

Historically, this technology has satisfied more than 90% of an engineer's flow applications and provided a fairly good accuracy.

It is commonplace, almost mandatory to verify the calibration of the dP transmitter.

First generation field instrumentation capable of transmitting a signal proportional to the process was mostly pneumatic. Unlike modern day electronic instrumentation, the mechanical force-balanced mechanism producing a pneumatic signal proportional to the flow rate required adjustments or calibration to keep them on track.

A technician typically performs this procedure, and it requires them to evacuate the process from the dP transmitter and into a container suitable for transporting the material for disposal.

Going back even farther, most processes simply vented to atmosphere. One should clear the tubing (decontaminated) assuring no product is left to migrate into the precision calibrator (pump up) before it is connected.

A procedure is available for each loop that lists special tools and Personal Protection Equipment (PPE). The more hazardous material in the process connection, the more precautions required. In the case of hydrogen cyanide (HC), PPE requirements include full containment suits with a breathing air source.

This proves to be an acceptable method of verifying calibration in non-lethal services, but for a small percentage of applications such as HC, sodium cyanide, and phosgene, there was no safe method.

In order to decontaminate a lethal service instrument, one must blow down the connecting tubing with nitrogen to a point of incineration or destruction.

Evacuating the cavity of a dP cell and tubing may leave residue in small concaves and seams in the connections. When the transmitter/tubing is broke down, the residue can produce air born particulates in many places exposing not only the technician, but also the personnel around him.

Other head producing devices are unacceptable. These devices have small cavities and narrow passages making decontamination difficult plus the risky procedure of extracting the device while the pipe is still in service.

Velocity meters such as magnetic meters have a liner that cyanide can easily permeate, and decontamination is impossible. Other meters like the Vortex Shedder lack the turndown need for maximum accuracy over a broad range or scale.

In addition, these lethal processes require reporting to state and/or federal entities in the event the process meets the ground or worse, personal exposure.

Some chemicals have harmful effects to humans in concentrations as small as 4 ppm, and there are stiff fines to ensure environmental compliance. All efforts to administer engineering controls over these lethal products are at work to assure personnel protection when working in the area.

One of these controls is the use of Coriolis flowmeters in place of head-producing devices. This minimizes exposure by reducing the number of leak points from 20+ down to only two being the flange connections. Since the Coriolis sensor has an inline flow path, there is less chance of the process being trapped in the piping.

Finally, since the Coriolis meter can read multivariable measurements, the density variable from the device can indicate a problem with the process without costly shutdown and pipe dismantling.

The Coriolis meter is mostly a flow through design that is self-cleaning and minimizes plugging. Normally, there would be no reason to decontaminate and remove the meter unless it was time to verify with a prover.

Due to the historic reliability of the Coriolis meter, the meters can go from one scheduled shutdown to another or further without the need to verify (prove) the meter.

Frequency estimates between proving vary depending on the application and performance desired. The meters will stay in service for many years without removal for proving and has shown in the past a remarkable retention of the K factor when it later went through verification.

This means the meter outperforms most units that require a shutdown every 18 months to three years.

Behind the byline

William Barnett (wbarnett@sterlingchemicals.com) is a project engineer at Sterling Chemicals. He is a longtime member of ISA and the president of the ISA Texas City Chapter. Rusty Liner (Rusty.Liner@EmersonProcess.com) is with Emerson Micro Motion.

A case of hydrocarbons and the lab

By Ralf Gusthuisen

Dual Coriolis meters replace pump and burette.

At a major hydrocarbon facility, a laboratory steam cracker works to investigate the products obtained by steam cracking of different feedstocks.

Feedstocks, which have not undergone cracking, must go through evaluation in an experimental lab furnace similar to the process operation conditions of the large-scale furnaces.

Databased models come out of these experimental results. Thus, the lab furnace is an important tool to improve the understanding of the steam cracking process and is a basis to optimize and control the production plants.

Due to the complex evaluation of the experiments, (different chromatographic analyses are necessary) and the high accuracy needed, it is important to ensure constant operation conditions when taking a sample of the produced raw gas.

There must be constant flow rates of the hydrocarbon and the water feed as well as a constant pressure and temperature at the outlet of the furnace.

Here is the experimental setup of the laboratory scale furnace before the mass flow controllers went online.

One major process condition is the feed rate of the hydrocarbons, which varies between 20 and 40 g/h. The corresponding feed rate of water is approximately 15 to 30 g/h.

A peristaltic dosing pump fed water. Feeding hydrocarbons by a pump is not possible as cavitation might occur. Hence, a pressurized burette fed the hydrocarbons.

The pressure difference between the furnace outlet and the burette is the driving force for the flow rate of the hydrocarbons. As no flow measurement was available, the most accurate way to calculate the flow rate was to measure the level in the burette at two different points of time and then to calculate the rate. If the flow rate differed from the set point, one adjusted the pressure in the burette.

Even though a good accuracy was possible using this approach, it had important drawbacks:

The time needed for the start-up of the plant depended strongly on the experience of the operator and usually took between 1 and 1.5 hours.

The flow rate of the hydrocarbons depends on the pressure difference between the burette pressure and the pressure at the outlet of the furnace. Certain feedstocks cause strong oscillations of the pressure at the outlet of the furnace due to significant formation of condensates in the tube. Thus, the feed rate of the hydrocarbons varied.

It was difficult to perform two experiments under the same process conditions. It was not possible to realize exactly the same flow rates in two independent experiments.

To overcome these drawbacks, automatic control of the feeds was necessary.

Setup of laboratory scale steam cracker before mass flow controllers were applied

  1. Holdup burette for hydrocarbons
  2. Holdup burette for water
  3. Dosing pump water
  4. Evaporator
  5. Furnace
  6. Condenser
    • Loupe to read off the level of burettes accurately
    • Manipulated variable to control the flow rate of hydrocarbons
Coriolis mass flow controllers refine and improve plant system

Novel lab-scale furnace

We installed two Brooks Instrument mass flow controllers to feed the hydrocarbons and water.

Because the plant runs a DCS (distributed control system), the measurements of the flow rate and the density are available online. In our application, 4-20 mA signals work. The set point also transmits by a 4-20 mA signal.

These are the advantages we realized by eliminating manual control.

  • Reduction of the time needed for the start-up of the plant
  • Improvement of the reproducibility of process conditions
  • Decrease of the number of hand sample that have to happen
  • Constant process conditions for the whole experiment, which is to say, the feed rates are independent of the pressure difference between the burettes and the outlet of the furnace.
  • Automatic control of the flow rates enables fast changes of process conditions.
  • Flow measurements are available online.

Now, the time needed for the start-up of the furnace does not depend on the feed rates but by the time needed to reach the desired furnace inlet temperature.

This takes about 30 minutes. In addition, the start-up time is nearly independent of the type of feedstock. In the past, the start-up of the process with a new feedstock took approximately 1.5 hours as the flow rates needed hand adjustment of the burette pressure (manipulated variable).

Indeed, before, after every change of the flow rate, the operator had to wait for the process to reach steady state before checking the flow rate once again.

Behind the byline

Ralf Gusthuisen works in the optimization cracker operations at Deutsche BP AG, Chemical Production Cologne in Germany.