1 September 2005

Heavy blow to overflow

Vibronic point level measurement for overfill protection.

By Dr. Christoph A. Rompf

You'll find liquids handling in all process facilities, including tank farms, food plants, chemical or pharmaceutical production sites, and water and wastewater industry facilities. Some of these liquids are toxic, flammable, and reactive, or they cause explosive gases and pose a risk to personnel and the environment. A facility operator has to assure these dangerous liquids stay in the appropriate pipes, tanks, and vessels. In particular, they need to avoid over-spilling a tank during filling processes.

Local laws, government regulations, pollution control agencies, and insurance companies require preventive measures to inhibit tank overruns (see Fig.1), especially during unattended automated filling processes. Irrespective of the federal and state regulations of a certain country, automated filling processes always require a high-level alarm that causes an automatic flow shut-off to prevent an overfill. The reliability and degree of functional safety of this overfill protection system are related to the potential danger of the liquid and the surrounding plant or facility.

Overfill protection systems

An overfill protection system should stop product flow during delivery before the tank becomes full and begins releasing liquid into the environment. As a general rule, such a system consists of a high-level sensor, a logic solver, and a final element that shuts off the flow into the tank. The system in Fig. 1B consists of a vibronic point level measurement device, an appropriate power supply, and a switch amplifier unit in the control room. A PLC- or DCS-based logic solver and a supply pipe shut-off valve complete the system.

The combination of these components has to fulfill the high functional safety demands defined in ANSI/ISA-84.00.01-2004, Functional Safety: Safety Instrumented Systems for the Process Industry Sector, or IEC 61508 and IEC 61511. High functional safety means these components either have to work reliably or give an alarm whenever maintaining the protection system. Thus, you must define different types of failures when discussing a safety system.

Overfilling a tank

Failures

A system's mean-time-between-failures (MTBF) generally characterizes that system or device. That number represents an average lifetime value for the system or device and includes all failures. With regard to a functional safety classification, not all failures are relevant. Failures the system detects and the alarm announces do not lead to critical situations. But failures leading to a malfunction and not detected and announced automatically are dangerous.

In the case of an overfill protection system, an operator would still rely on the system. In case of a demand, the system would fail. To track down these dangerous failures and to quantify the likelihood of a dangerous failure on demand, do a failure mode, effect, and diagnostics analysis (FMEDA). The guidelines for this appear in the standards IEC 61508/IEC 61511 and ANSI/ISA-84.00.01-2004. Using these guidelines, you can calculate a probability of a dangerous failure on demand (PFD). According to the different safety integrity levels (SIL1-SIL4), the probability of a dangerous failure cannot exceed given values.

To make the determination of a SIL classification manageable, divide the safety considerations into different components. Split the PFD values into the different components. A general recommendation is to weight the PFD value 35% on the sensor system, 15% on the logic solver, and 50% on the final element (see Fig.2). Finally, review the whole system. Account for statistical failures, and avoid systematical dangerous failures.

Point level measurement

Point level measuring devices for liquids see use in all process facilities. Numerous measurement technologies are available for these kinds of applications, such as float switches, vibration limit switches, ultrasonic gap switches, and capacitive or conductive limit switches. When considering high functional safety, choose a measurement method with low PFD values and no systematical failures during operation. Also, do a thorough investigation of the measurement sensor systematical failures, and consider the know-how gained from a large number of applications. Also, do a proven-in-use evaluation.

Vibronic point level

Vibronic measurement devices or tuning-fork systems fulfill the highest demands with regard to safety and reliability. The main advantage of vibronic point-level measurement devices over float switches with regard to functional safety is they use an active measurement principle. The device stays in vibration continuously, and an evaluation electronics always monitors it. It immediately detects sensor failure in almost all cases and avoids dangerous failures. Vibronic measurement devices meet the requirements of almost all point-level applications for liquids. A vibronic device is independent of the installation position. You can mount the same device from the top, the side, or the bottom. State-of-the-art devices have a broad application band-width where no calibration of the sensor is necessary. This sensor is independent of process influences, such as pressure and temperature. It's independent of substance characteristics, such as conductivity, dielectric constant, and viscosity, and is independent of gas bubbles, foam, and solids in the process liquid.

These systems are gaining acceptance as a standard solution for level limit detection in all industries and work properly in a wide range of applications. You can determine functional safety data from this high number of installed units and make proven-in-use evaluations. One example is a tuning fork system installed in more than 1.5 million applications, leading to a sensor design, optimized for overfill protection systems, and meeting the SIL2 level in a one-out-of-one (1oo1) and SIL3 level in a 1oo2 or 2oo3 installation architecture.

Distribution of probability of failure on demand

Vibration limit switch operating principle

Mechanically excitable systems see use as vibration limit switches. These are usually oscillating forks with two tines, which a piezo drive excites and converts electrical energy into mechanical energy. A second piezo acts as a receiver re-converting the mechanical energy into an electric signal again. This electrical signal amplifies, phase shifts, amplifies a second time, and feeds the piezo drive. Thus, an electro mechanic loop sets up and acts as a basic wave excitation and always causes the tines to oscillate with their resonance frequency. The setup of this basic wave excitation appears in Figure 3A.

Fig. 3: Vibronic measurement devices - A: Basic wave excitation; B: Frequency-immersion

Characteristics

Liquid surrounding the tines extends the mass of the resonance system. The frequency reduces by the immersion of the tines in a liquid. Evaluation electronics monitor this frequency shift. Below a certain frequency, the sensor reports the covered condition to the evaluation electronics, which indicates the switch point. See the typical characteristics showing the dependence of the resonance frequency of the depth of immersion in Figure 3B.

A switch hysteresis of nearly 30 Hz sees use between the activation and deactivation point (fE and fA in Fig.3) to reduce sensitivity to state changes. This corresponds to a hysteresis of the switch point of approximately 0.1 in. Furthermore, a time delay of about 1 second prevents a strong dependence on turbulent currents and waves on the surface of the measured liquid.

Corrosion

In general, it's critical to consider sensor corrosion when applying point level measurement devices for overfill protection. If you manage corrosion improperly, you could face a state of dangerous failure. You can't ensure the safe operation of a passive sensor because corrosion is impossible to detect. In contrast, a sensor with a high functional safety has to operate properly, even if it is partly corroded, or you have to report the function failure to the control system. Vibronic point level measurement devices meet this demand.

Behind the Byline

Dr. Christoph A. Rompf is a product manager for level measurement at Endress+Hauser GmbH + Co. KG in Maulburg, Germany.

Ultrasonics vs. guided-wave radar

By Ellen Fussell Policastro

The big advantage of ultrasonics, and one of the things making the technology more reliable, is the rising popularity of microcontrollers said Karl Reid , product line manager at STI in Logan, Utah. "With microcontrollers, you're able to make smarter probes that will allow you to filter out some of the anomalies and inconsistencies. We have to see a target a certain number of times before we say it's a real target. We can filter out electrical or acoustic noise. We can put in windows around problematic objects so the reading ignores those objects."

When measuring level in a water tank, the easiest place to mount the sensor is close to a ladder or a pipe. "With ultrasonics, we can filter out the pipe or rungs of ladder and focus on the level," Reid said. "Even if we're getting reflections from other objects, we can ignore them."

It's much more difficult to use a float level in hazardous environments, Reid said, because the float must touch the liquid. "If the liquid has a tendency to build up, the float can stick. There's constant maintenance because it's down in liquid. Some liquids in oil fields have caustic mixtures. The advantage of ultrasonics is it doesn't touch the material. There are fewer maintenance issues, and it's easier to install.

Even installing resistive chain sensors with probes up to 20-feet long is difficult, he said. These large reed switches in magnets, activated when the sensor goes inside the magnetic field, are tied into resistor. As the float moves up and down, it will close the resistors at different points and give a different voltage output, depending on which is closed.

And just like all measurement technologies, level is paying its dues in the environmental, security, and standard compliance arena. Reid's been working with a group monitoring level in trucks transporting diesel fuel. "They seal it up, but they look at the level of the diesel staying consistent within a certain window while it's being transported," he said. Reid is also seeing more monitoring of level in different tanks to avoid spilling and to detect leaks. One of the biggest new standards his customers have to follow deals with oil fields and environmental standards for leak detection. "Companies drilling for oil have to account for the difference in how much they put in and how much comes out when drilling into the ground," he said.

Reflections on radar

Chris Romano, a product manager at Pepperl+Fusch in Twinsburg, Ohio, deals with guided-wave and free-standing radar systems, using a technology called time domain reflectometry (TDR), a new way to measure high-frequency signals. The method, originating from Lawrence Livermore Labs in Berkley, Calif., sends high pulse signals to reflect off a surface and come back.

"The key is it's high frequency, very guided, and not susceptible to temperature changes, dust, humidity, pressure, or changes in the medium (physical or electrical properties)," Romano said. Boasting durability in the industrial market, TDR also has experience in detecting rebar (metal joints) inside concrete and in metal detection. Mercedes is looking at using TDR in collision control because it senses through plastics, water, and humidity.

Guided wave radar sends a signal through a cable or a rod, Romano said. "When it runs down the cable and it hits the medium (because it has a different bioelectric than air), it reflects and sends the signal back up through the rod. The amount of time it takes will tell us the distance you are from the medium."

Free-standing radar uses an antenna that sticks down so far, depending on applications. The signal travels through air or inside a tank using the same principle. "Except instead of using the rod or cable as your medium to send a signal, you're using air," he said. The signal reflects off the surface, and then comes back to the antenna. Again, the amount of time it takes to send and receive will tell you the distance and thus the level.

There are other types of non-contact level probes, such as ultrasonic. "But the problem is, if you get dust, humidity, water, or foam on the surface, you get a bad signal," Romano said. Guided-wave radar is immune to those problems. "However it is still susceptible to foam, dust, and water in atmosphere, not as much as ultrasonics, but they still have problems."

With guided wave radar, "you're not using the air as your medium to send a signal, you're using the cable or the rod. Because of that you could care less about the humidity, dust, or anything else. You're using the probe or the cable or rod as your mode of sending the signal. Guided wave radar is better, but you have to make contact," Romano said. Free-standing is true non-contact and is useful in measuring level when there's a substance, such as concrete, rotating in the bottom of a tank.

Coming Next

Ask The Experts

Read questions answered by our experts or join the email list.